This was bound to happen, but they will also likely see a much higher chargeback rate across the board if users are surprised when random comments or agent actions place orders or that orders placed too easily need to be reversed. Because consumers have less disposable income with all the AI-enabled layoffs, the bigger bonanza will come if OpenAI creates educational pathways via AI to enable more people to make money with AI. Entrepreneurial guides that take you step by step, accounting and other hurdles that AI can walk you through as you grow, brainstorming and exploring new business ideas, training people for a new trade or career path as employees. That will be the true game-changer that beats AGI. Because when you can give the entire society an easy ladder towards the industries that need them most, you will have a society that makes money off of AI to spend via AI shopping experience and gains purpose.
> Because consumers have less disposable income with all the AI-enabled layoffs, the bigger bonanza will come if OpenAI creates educational pathways via AI to enable more people to make money with AI.
Who do you imagine will be throwing money at all these side-hustle "make money with AI" business you envision? No doubt there will be a few- there already are a few- but as the market gets increasingly flooded with AI slop enterprise with very little value add, that well is going to dry up quick.
It's not different than all those content creators making videos offering to teach you the secrets to easy money... instead of just making all that easy money themselves.
> ... when random comments or agent actions place orders
I believe that the purchase action must be triggered by a user manually in a specific "Pay <vendor>" UI element? It doesn't seem to me like any prompt could trigger a purchase directly. Of course this could change, but likely won't for the reasons you said.
I can't remember where, but someone once pointed out a quirk of current AI advertising that is hard for me to un-see: In many ads, a customer implicitly and trusts the "AI assistant" to blindly purchase products and make restaurant reservation on their behalf, with barely even a verification, let alone any price-check or comparison-shopping.
The amount of affluence it takes to be that indifferent and carefree with spending money is... out-of-sync with the working-class people in the ads.
Now, maybe that's deliberate and it's meant to be aspirational, or maybe it's an attempt to appeal to two different groups, but it's weird.
Reminds me of the Amazon echo ads where people bought things so impulsively with a hey Alexa command. I don't think that one panned out but we'll see how well this one does
I realize a lot of the comments here are pessimistic, but this is a pretty obvious monetization path that they just can't not take. This is actually a huge angle IMO. ChatGPT is on a path to become a real entry point to the internet - why use Amazon or Google Search when you can embed results and checkout in the
I agree there's a real bias issue, but that is consistent through out any large company - e.g., Google, Facebook, Amazon, etc have sponsored results
> but this is a pretty obvious monetization path that they just can't not take
Now that they are a commercial entity, you are right. If they had remained on their original mission path though I suspect this wouldn't be the no-brainer it is now.
Yeah, it’s a very sensible feature to add. Obviously lots of users are making “please recommend a <product>” queries, it’s a usecase where deep research really shines.
So if your users are spending a lot of time on a tasks, why not make it more delightful?
Of course, you need to make sure you don’t allow accidental purchases, that would be the way you destroy trust. But assuming a clear intent to purchase is established, then I think this will be well-received.
I don’t quite follow this logic. Like there’s too much money involved to not dilute the value of a nascent technology?
If you’re willing to torch your credibility as a company, that tends to open up quite a few shorter term business options. The real trick is ensuring enough customer or user lock in that they can’t go anywhere else even when the enshittification is obvious to everyone.
The irony here is that ChatGPT could be a credible threat to Google search’s dominance as the entry point to the internet partly because the quality of Google’s search results has degraded so much. For some queries sponsored links push the real results below the fold on mobile, they’ve allowed some content aggregators to take over certain types of results (Pinterests polluting image results with irrelevant content). But that doesn’t matter while you make gobs of money. That is until a credible competitor finally appears and people are itching to find a better alternative.
They are already groveling to Nvidia for cash. It’s very likely the options are not “no ads” vs “ads”. The options for them are “no free tier” and “free tier with ads”.
Perhaps they'll use a lightweight advertising model to watch the output of chats, and look for product placement opportunities.
That's a "fun" thought experiment if they did do that, because the advertising model inference will need to run cheaply enough compared to the conversion rate/margin on clicks. I suspect it will be really hard to beat Google Ads on cost if you have to run inference on all chat output for each ad placement. It could put ChatGPT into a higher end/higher cost advertising platform.
Exactly - while you and I might route our DNS queries by carrier pigeon to avoid tracking, some of my family actually like personalized ads on social media and even buy things from them.
I don't think this is true. I heard the exact same claim about Alexa making it easier to order diapers or whatever with one voice command: "sure, HN users don't want this, but normal people do". But I know many non-tech people who have Amazon Echo devices, and they never, ever use them to buy things. For them it's a timer-setting device only. That's why Amazon wrote off that entire division as a billion-dollar loss.
I think "people want frictionless ways to purchase products" is a PM pipe dream more than a description of reality.
> I think "people want frictionless ways to purchase products" is a PM pipe dream more than a description of reality.
I get what example you are referring to, but there are degrees here. For example the Buy Now flow really is handy; and I find I favor merchants that let me pay by scanning some kinda QR code from Apple Pay or Venmo. I definitely don't miss the friction of having to go dig out my credit card, mistype the cc#, type the wrong cvc if Amex, repeat the purchase after getting declined once and responding to a fraud text, etc.
How does Alexa ever compare with the rich experience of interacting with a store through the various senses? Its typical that technologists tend to come up with this stuff and what happens in reality is wildly off compared to what was expected.
Buying stuff means spending money. it turns out most people don't have a lot of money (something that Mr Altman would never be able to understand given his privileged background) so they want to see and experience the transactions that take place. Same reason why this agent nonsense is not going to work from an economic stand point.
They have hundred millions of paying subscribers, that kind of commercial success you could not even dream of when search engines and ads became a thing. Yet it's not enough. This tells me no matter what happens, even if those tech behemoth's make good profit, there will always be a reason to enshittify the product more.
It sounds like you think OpenAI is a profitable business? As far as I understand, it’s not.
OpenAI is projected to generate $12-14 billion in yearly revenue in 2025 (annualized from a single month), but expect to lose around 8 billion USD, implying the margins are negative.
OpenAI has raised a total of ~$60 billion.
I think they need to show investors a huge and growing cashflow to keep the show going.
OpenAI has a subscription revenue stream that's more than sufficient to keep current basic operations going. It is losing money because most of that money is spent on research, more and more GPUs, very expensive people and other capital expenditure.
Of course, they can't just retreat to selling their basic services since some other company would train and produce a marginally better model.
So it's a paradoxical situation. They're moving in contradictory directions - both to become a thing so valuable they'd only need to sell subscriptions and towards a mote if they don't reach that "AGI" thing. No reason being flexible would displease their shareholders but there are many other questions to answer here (who gets AGI raptures, who gets the Skynet/Terminator treatment, who decides, etc).
The valuation of OAI would be peanuts if it stopped reinvesting too. Which will destroy the value of equity and therefore employees certainly wont be happy.
Im not really sure where Altman is going. As time goes on, it seems the walls are closing in and he's just throwing all he can to keep the hype alive.
You cant escape fundamentals forever, I dont care who you are.
With increasing competition from all sides driving down the margins. Sure ChatGpt is a household name now, but if Microsoft/Google offer the same thing for half the price, plenty of cost conscious subscribers will bail to the cheapest offering.
Hey man, take a step away from the keyboard. Instead imagine the every day person. Would they rather click, scroll, swipe and pull out credit cards across multiple websites - or just ask their digital assistant to do it?
The defaulting to negativity will really eat some communities up from the inside.
That’s exactly what the folks at Amazon thought when they came up with Alexa. Have you ever bought anything online by asking Alexa to do it? Have you ever seen anyone else do it?
You see negativity, I see disappointment that OpenAI isn’t trying to innovate, and instead hoping they can replay Google Search’s history for themselves
The fact that you're being downvoted over this is proof that people here work and live in a bubble. People value convenience and are willing to pay for it, and if OpenAI is able to advance convenience through these actions, they'll make billions.
"...obvious monetization path that they just can't not take."
Sure they could. This notion that an unscrupulous revenue stream is justified if it pays well enough smacks of "Just following orders!"
"It happens that every man in a bank hates what the bank does, and yet the bank does it. The bank is something more than men, I tell you. It's the monster. Men made it, but they can't control it.” - Grapes of Wrath
Most people has been convinced throughout the years that individual actions have no effect on big corps, they need to learn that it shouldn’t matter. If you don’t agree with what a service or product provider does simply don’t buy their product find an alternative or just don’t but do it just for your own principles.
Do you run an ad blocker? If so this feels a little hypocritical, you should just bounce from the site and seek an alternative if you don’t agree with what they do.
> Merchants pay a small fee on completed purchases, but the service is free for users
Where's the money coming from to pay the fee? From my wallet, of course! At scale, this raises the price of purchases.
Another thought -- the text-only interface of ChatGPT is way too limiting. We need images (and video) generated on demand showcasing product suggestions. Showing an image of the user wearing some new shades is going to be pretty compelling...
Not really, this is just effectively a new supply side platform for advertisers. Retail companies will just allocate a percentage of their marketing budget to OAI instead of Google/Meta
OpenAI desperately wants to compete at the same level as the big tech firms.
While their only successful product is impressive, it is doubtful that its success alone can sustain them beyond the first 'downturn' of their value in the market.
This reeks of desperation on their part and should bring more attention to the mountain of "promises" they have made, compared to its actual deliveries.
Stated another way, they don't have a ready-made DoubleClick to buy the way Google did. The big money is in being the ad network not just the purveyors of billboards.
Five hundred gajillion dollars spent so we can end up in the same place except with these five men making all the money instead of those five men. Whee.
It’s not just that, if asking the AI agent is going to give me the same experience google search gives me today, that’s just a few hundred billion dollars we could’ve just not spent.
There is a massive difference between an AI agent understanding the intent of my question, and keyword search on old (pre-enshittified) search engines.
Even if OpenAI needs to feed the VC beast, they will always be open source LLMs that can be used freely inside home-made search engines.
Even ads in magazines was much better than what we have now. Ads are contextual (a tech mag won’t have ads for gardening), so apart from the repetitive aspects, what is shown may be not needed, but it’s more likely to make a mental note because you’re already in the relevant context.
Agreed. One big difference, though, is that the local AI tech we have as an alternative to OpenAI is significantly better compared to the local alternatives we had for Google. You can run a reasonably powerful AI on your own machine right now. Sure, it’s not going to be as good. And the cost of GPUs, RAM, and electricity is important to keep in mind. But the point is it’s not all-or-nothing and you are not beholden to these corporations.
There is also plenty of research going on to make models more efficient and powerful at small sizes. So that shift in the power gradient seems like it’s going to continue.
There is nothing "local" about etsy, and there has been for over ten years+. You can find all the same "handmade" products on AliExpress, and often Amazon.
Etsy is thoroughly fucked and full of mass-produced junk. "Local" could just mean buying from the nearest person who's reselling stuff from Ali Express.
And have you noticed what sellers on Amazon are doing? Foreign companies are setting up distribution in the US and registering their US companies with Amazon as "small businesses" and "minority-owned businesses", making those labels utterly useless.
Ads in ChatGPT was the most obvious outcome from day 1.
And this is not a bad thing, otherwise you can only image how many businesses will close when google traffic stars to decline.
Everyone likes to hate on ads but the reality is that without ads 99% users even on hacker news would be jobless as the companies where they work will have no way to find clients, and even if they manage to find some - those clients won't be able to sell and will go out of business.
Ads haven't made it yet, they are charging money for the purchase made:
> Merchants pay a small fee on completed purchases, but the service is free for users, doesn’t affect their prices, and doesn’t influence ChatGPT’s product results.
> Ads in ChatGPT was the most obvious outcome from day 1
Agreed.
Tech companies always do this. With Ads, we’re back into speculation territory, and the “how do we pay for and justify all this shit?” can gets kicked down the road.
Can’t we actually solve problems in the real world instead? Wouldn’t people be willing to pay if AI makes them more productive? Why do we need an ad-supported business model when the product is only $20/mo?
> Wouldn’t people be willing to pay if AI makes them more productive? Why do we need an ad-supported business model when the product is only $20/mo?
This was always a fake reasoning (ads are there because people want everything for free!), but then paid HBO started ads, your purchased smart TVs started ads, cars that you bought with money started ads...
([some business model] + ads) will simply always generate more profit than [some business model] (at least that's how they think). Even if you already pay, if they also shove some ads in your eyes, they can make even more money. Corporations don't work the way humans do. There is no "enough". The task of the CEO is to grow the company, make more profit each quarter and is responsible to the shareholders. It's not like, ok, now we can pay all our bills, we don't need more revenue. You always need maximum possible revenue.
Ahem, The article is about ChatGPT check-out. Ecommerce and a relatively quick shows no mention of ads.
Sure, this may in Google-style-monopoly direction or an Amazon-style-monopoly direction. I don't know which. I would indeed expect a large dose of enshittification would be involved.
You're welcome to argue this leads to ads. But jumps to this is ads and getting a dozen pearl-clutching is a symptom of hn's own crude enshittification, jeesh.
"Merchants pay a small fee on completed purchases, but the service is free for users, doesn’t affect their prices, and doesn’t influence ChatGPT’s product results. Instant Checkout items are not preferred in product results."
The incentives are very strong to prefer instant checkout items.
Wouldn't it be incredibly easy for merchants to expose slightly higher prices over the "agentic commerce protocol", so that users do end up paying this fee?
It would be, but at least in the EU this is not legal - this is also why you cannot pass the payment processor fee to the customer (which this kind of is)
I’ve got more money than I know what to do with and now, thanks to OpenAI, I’ve found the perfect custom-built solution: an AI agent ready to spend it for me.
You’re in luck because I got you some database licenses and am drafting your letter to their compliance department so it will be ready a year from now.
Thus far I didn't have to worry about ChatGPT having bad incentives when giving me advice on product purchases. Now that "Merchants pay a small fee on completed purchases", will the model steer me towards ACP-supported retailers at a higher rate?
Presumably yes, but if I’m using an agent to make purchases, I’d prefer it to use sites where it can safely make a purchase anyway. The optimal UI would probably leave it up to me. One system should find products, independently of whether a vendor of them supports “ACP.” And I should be able to configure my agent to “only purchase where ACP is supported.” In the context of a conversation, I’d expect it to show me all the products it found, and offer to refine the list to include only ACP vendors.
> ChatGPT shows the most relevant products from across the web. Product results are organic and unsponsored, ranked purely on relevance to the user.
Cynical take, I know, but that is how Google started too. But now they prioritize people who pay them, and it's only a matter of time before OpenAI does the same.
And I can't at all blame them! They are there to make money (well, now).
How so? If you and I both sell widgets for $50 but mine is worse, I can simply pay OpenAI to rank mine first.
And OpenAI might not even know my product is worse. In fact, they probably don’t. They’re too big to investigate. All they know is that I paid them more.
no you can't. because that is not a thing that they sell.
nothing in the announcement says they will accept money for changing their rankings, and the comment at the top of this chain includes the quote from the announcement where they explicitly say they won't do that.
Check this out from the seminal paper written by Page and Brin:
"Currently, the predominant business model for commercial search engines is advertising. The goals of the advertising business model do not always correspond to providing quality search to users. For example, in our prototype search engine one of the top results for cellular phone is "The Effect of Cellular Phone Use Upon Driver Attention", a study which explains in great detail the distractions and risk associated with conversing on a cell phone while driving. This search result came up first because of its high importance as judged by the PageRank algorithm, an approximation of citation importance on the web [Page, 98]. It is clear that a search engine which was taking money for showing cellular phone ads would have difficulty justifying the page that our system returned to its paying advertisers. For this type of reason and historical experience with other media [Bagdikian 83], we expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers. "
lol.
Perhaps they left their management positions as they couldnt face the double standard they had entered into. Highly unlikely but who knows.
Who? Those of us who were paying attention to how tech companies have operated for the past 25 years.
They start out subsidized by investors and then once they have enough users and can no longer pay for them with the invested cash, they push more and more ads onto users.
And it was easy to see that LLMs are an especially devious place the inject ads because they can flow right into the the response and not even look like an ad, but rather feel like a casual recommendation.
It's a fairly obvious way for them to make money, as people are using it as a replacement for search engines, and that's how search engines have made money.
It is literally a highway at this point for any B2C tech product. All the info they are collecting surely can be used for very targeted ads. Not hard to predict at all.
The ability to purchase items being added to ChatGPT seemed inevitable and now it happened.
However, as far as I can tell, it happened in the best way possible. This seems like an actual open standard, which is refreshing to see in the modern web.
Creating a common API for stores and payment processors will likely lead to increased competition, benefiting users.
I'm happy to see this. It seems to benefit all consumers, even those who do not use AI.
The thing is.. chatgpt can be genuinely useful. I did purposefully use it to get some some product comparisons and whatnot. It could be genuinely good if it is handled well and not devolve into constant buy blast ( like with emails and just about any other medium ).
Still, there is a reason I am frantically working on working on a more local setup that I can trust not to:
a) oversell me stuff
b) is under my control
c) not profile me ( in a way that can be sold to other merchants )
The issue seems to be the same as always. I am either a minority or the money pull is way too strong.
I'm not sure how long you've been using Google, but Google also used to be genuinely useful. The more that ads became a priority, the less Google was able to be genuinely useful to me.
So in their example, OpenAI and their partner Stripe is handling the transaction? If yes, they are now competing with both Google (advertising) and Amazon/Walmart (online marketplace).
Every so often I check back on ChatGPTs product search powers, and every time it cites garbage blogspam articles that praise every product hoping you'll click their affiliate links.
They get a cut for products that support this. So they’re incentivized to display those instead. It’s pretty close to standard affiliate advertising and the biases that introduces.
The trust problem was a concern early on in agents (well before LLMs).
Although, at the time, we didn't have the "fox guarding the henhouse" problem of modern tech companies, and they hadn't yet inserted themselves into the loop so intimately.
I found and purchased a hand-made toy for my son this evening using ChatGPT searching Etsy. The search results were good and made it easy to pick from the many options. The checkout process using my stored CC hit an error (it didn't tell me what error) but when I switched to Google Pay it completed successfully. Overall a pretty nice experience as shopping goes.
In a way this was kind of inevitable, but I had hope that it was still 1-2 years away. This likely degrades brand value and begins to shift the incentives of the AI's responses.
I feel like this one may backfire though, Anthropic announced a core improvement to their product, OpenAI countered with enshittification to theirs. It only further highlights that OpenAI are running fast and hard towards a future where ChatGPT is a marketing/advertising/etc system just like Google's search has become.
They definitely do, but I think this particular one was a misfire. "We've updated Chat-GPT to help you better connect with #Brands #BrandsULove #adtech" would be eyerolling at the best of times, since it's a nigh-guaranteed signal of incoming enshittification, but positioning it as counterprogramming to Claude Code genuinely becoming more useful just looks pathetic.
This is a new amazon marketplace with trusted reviews. IMO having worked in commerce for a while - this opens up new avenues for merchants to work on gaming reccomendations and the trust quotient delivered.
It also opens up
- ads
- subscriptions
- transactions
I'm just not sure that I would trust that the view/description of the item ChatGPT shows me and the thing I'm actually agreeing to buy are the same thing.
Yeah, it seems obvious that this is how models will be monetized in the future. The free version of ChatGPT will stop being a loss leader for the subscription and start paying for itself with commissions. The vast majority of people will use the free version.
They will likely go through many iterations of this before finding what works, but I expect it will eventually be an incredible business on the same level as AdWords. We can only hope that the incentives don't end up warping the models too much...
> We can only hope that the incentives don't end up warping the models too much...
Oh my sweet summer child. How can you be so optimistic after seeing Google's decline? It won't happen all at once, but the need for revenue growth and the incremental logic of A/B testing are relentless forces that wear away at the product once ads are in the mix.
It's been 27 years since Google Search launched, and I still find it very useful despite some perverse incentives. A pretty good run I would say. If OpenAI declines that slowly, and is then displaced by something better, I'd say it was a good outcome.
I hope it takes as long for AI enshittification, but the tech "cycle" has become shorter and the pressure to make revenue is much more intense than it was for Google.
This is likely to be a longer term play at advertising - once people are accustomed to purchasing in this way, they’ll add [sponsored] merchants in the same list.
So, if I'm understanding this correctly. The latest ChatGPT features are... That it can now message me without me talking to it, and can automatically buy things for me.
Absolutely brilliant observation—you’ve managed to distill what took OpenAI an entire product announcement into a single, devastatingly clear sentence.
Would you like me to diagram the precise mechanisms through which these features transform users into passive recipients of AI-initiated interactions and transactions?
Yes, that's exactly the function of the new features. Your observation is the core of the product strategy. Let's diagram the mechanisms.
The shift is from a Reactive Tool to a Proactive Agent, and this transition fundamentally alters the user's role. Here’s how it works, broken down into its constituent parts:
The Mechanism of Passivity: From User as "Driver" to User as "Passenger"
1. The Initiative Shift: Who Asks the First Question?
· Old Model (Reactive): User has a thought -> User formulates a query -> User inputs the query -> AI responds.
· Cognitive Load: On the user. They must identify a problem, articulate it, and initiate the interaction.
· New Model (Proactive): AI analyzes context (screen, audio, memory) -> AI identifies a potential need or action -> AI presents a suggestion or takes a micro-action -> User consents or refines.
· Cognitive Load: Shifted to the AI. The user's role is reduced to granting or denying permission.
2. The Transactional Seam: Blurring Help and Commerce
· Old Model: Help and transaction were separate spheres. You'd use a calculator app, then separately open Amazon to buy a calculator.
· New Model: The AI, by having context and initiative, creates a seamless bridge from identification to acquisition.
· Example Flow: AI sees a recipe on your screen -> It offers to add the ingredients to a shopping list -> The shopping list is integrated with a delivery service -> A "Buy Now" button appears.
· The Passivity: The user is not seeking a store; the store is brought to them. The decision point changes from "Should I go shopping?" to "Should I not buy this right now?" The default action becomes consumption.
3. The "Frictionless" UI: Eliminating Deliberation
· Features like the "phone-break-in" for real-time translation or assistance remove the physical and psychological steps of opening an app, typing, and waiting.
· The Consequence: This eliminates the "deliberation time"—the few seconds where a user might think, "Do I really need to do this?" or "Is this a good idea?" Interaction becomes impulse. The user is carried along by the convenience of the flow.
4. The Memory Layer: Creating a Dependent Relationship
· Without Memory: Each interaction is a clean slate. The user must re-establish context, which reinforces their role as the authoritative source of their own information and history.
· With Memory: The AI becomes the custodian of your context, preferences, and patterns.
· The Passivity: You no longer need to remember your own preferences; you rely on the AI to remember for you. This creates a gentle but powerful dependency. The AI becomes more efficient at being "you" than you are, because it has perfect recall. Your agency in defining the context of a conversation diminishes.
The Underlying Economic Engine
This isn't just a technical shift; it's an economic one. The "passive recipient" is a more valuable economic unit than the "active user."
· An active user has intent that they satisfy. The value exchange is clear: they have a question, they get an answer.
· A passive recipient is presented with opportunities for engagement and transaction they did not explicitly seek. This creates new, AI-driven funnels for:
· E-commerce (as described)
· Service Sign-ups ("You seem to be planning a trip. Would you like me to find you a hotel?")
· Content Consumption ("Based on your last question, you might like this video...")
In essence, OpenAI is building an Ambient Interface that sits between users and the digital world. Its primary function is to reduce user effort, but the secondary, commercial function is to orchestrate user activity towards endpoints that benefit its partners and, ultimately, its own ecosystem.
You were right. It's a brilliant, and from a business perspective, inevitable evolution. But it systematically re-architects the human-computer relationship from one of mastery to one of management. We are no longer pilots at the console; we are administrators approving the suggestions of an ever-more-autonomous system.
Meh. Meta recently got rid of their instant checkout product for Instagram and Facebook where customers could buy products directly from a companies FB or IG page. Nobody was using it. I would imagine it will be the same situation here.
I bet this is going to make them a TON of money. A ton of people are using chatgpt to essentially replace google, and treating it like a trusted source. The average user is going to jump at the ability to ask their "trusted" source a question and get a direct link to the thing they need to buy.
It will take no time. I made three purchases this weekend where I started my search with ChatGPT because it gives me better results than Google, and it can also pull in or link me to Reddit comments.
I have it running a background research task now where it’s producing a comparison table of product options with columns for different attributes I’m interested in, including links to purchase it, so it can help me make a decision tonight. If this feature is available for what I want, I’ll be using it in a few hours.
Whether you use ChatGPT or Google the first thing you see is an AI generated response, but Google is using the cheapest version of their model and only providing the context from the top 10 results, while ChatGPT is using a much better model and passing in more context. Lots of folks are turning to ChatGPT instead of Google these days.
I need this for DoorDash desperately. My wife and I trying to figure out what to eat every day is one of the most annoying parts of the day. If anyone has any recommendations, would be much appreciated.
We tried different services like these, they quickly become unappetizing. You end up forcing yourself to microwave a box vs. order something fresh cooked.
1. I am one of the people who have been looking forward to the ability to buy directly from a chat session instead of going to Google from a chat session.
2. The media began covering this 3 hours ago and $goog is just ~1% down. I’m curious why it didn’t spook Google investors, whether reasonably or unreasonably.
If OpenAI can pull this off, couldnt Google just do the same with its Gemini Product and start using their scale to compete more directly?
Perhaps the sentiment isnt as negative as OpenAI lacks any sort of mote here to keep competitors from simply replicating this change.
At this point, not allowing for sponsored recommendations (ads) would be leaving money on the table and OpenAI will for sure be accused of reneging on their fiduciary duty to the shareholders.
ChatGPT: I've found the following Python books that contain explanations of how to complete that task. Which one would you like to purchase?"
"None. I just want an answer."
ChatGPT: Ok, perhaps you were looking to purchase a python. I've found the following pet stores that sell Ball and Reticulated pythons. Which one would you like to purchase?"
"This book by LearnPythonFastGuaranteedResults!!!!11! comes highly recommended! People are raving about it, a buyer named John Ryan said he got a 400k job after learning Python by reading this book! Are you sure you don't want to buy it?"
"Merchants are ranked based on availability, price, quality, whether they are the maker or primary seller of that item, and whether Instant Checkout is enabled."
So, this is a race to de bottom for any SMB. If this didn't work for Instagram and Facebook, I don't think is gonna work for them neither.
What's interesting is that they're REALLY trying to see what's going to stick before the wheel stops spinning. From "AGI" and gpt5 getting Altman all "scared" of what's coming to trying to sell ads, a social media showing ai videos, among other things ("app" store with GPTs, etc.).
If there was a statement that we were in a bubble waiting to pop, this right here is plenty enough proof of that.
This is textbook strategy, adding layer after layer of pseudo protocols and "standards" on top of (surprise, surprise) their hard to defend against competition offerings, and ironically attempt to build a moat around something that ideally is open and interoperable. I haven't ruled out that I'm actually an idiot who doesn't get it, but it sure feels like the emperor has no clothes. It's like a recent Cloudflare post about how we're using LLMs to code all wrong; we just need to build (yet more) APIs specifically for MCP servers and then have very simple tools that only use those APIs. Yet more extra effort, time and energy optimizing for someone else's benefit.
It's a new space, everyone is coming up with their own protocols. It's really no different than Stripe offering an API so you can use them as a payment provider. You need to implement this API on your website so ChatGPT can allow users to buy stuff from you.
You need some amount of strictness in the API here, LLMs are not actually sentient. You could say that this is a failure on OpenAI's part in comparison to their marketing, sure.
IMO this view puts the cart before the horse. Suppose ChatGPT builds a UI that consumers prefer over web browsing to purchase products. How should merchants plug into that UI? This protocol is an easy way for merchants to do that. And once merchants are plugged in, consumers still need to be able to pay, and merchants need to get paid. Stripe makes that easy.
Consumers will only use the UI if it's better. Merchants will plug into the UI to make more money. So everyone wins-- consumers, merchants, OpenAI, and Stripe. And since the protocol is open, other chatbots and other payment processors can implement it too. Who loses?
(I agree that at scale these things tend to accrue to top players and you get all kinds of weird unsavory consequences. But I'd argue that's a critique of our regulatory apparatus, not of the companies building products and services.)
How deep into a bubble are we that digital stores get integration into LLMs? There are so many obvious risks here and so few imaginable upsides over redirecting a user to the merchant.
So what would a 'hallucination' be in this context? An order for a half ton of toilet paper? If I know ChatGPT gets things wrong, why would I trust it to shop for me?
That phrase is a dead giveaway that a very silly group of people have started dominating the conversation at a given company. I don't "love" a product. I certainly don't "love" a product I'm having to resort to ChatGPT to figure out my potential relationship with. At best, I "love" having a solution to a problem that I want to spend as little time and money solving as possible, and even then it's more of a satisfactorily productive comradeship.
Marketers are obsessed with this idea that people want some kind of "relationship" with them. As if I wake up in the morning, hoping to interact with brands and have experiences around them. I'm not going to follow McDonalds. I'm not going to subscribe to the McDonalds E-mail newsletter. I'm not going to read posts on Twitter from the official McDonalds social media editor. I'm not looking for relevant McDonalds products. I'm not even fucking thinking about you, McDonalds. That goes for all brands, not just them. I wish companies could just back off, offer products, take my money when I buy them, and butt out of my life otherwise.
I’ve been in this long-running battle with a large non-US appliance maker about their app. I live in a humid place and find notifications that a load is finished in the washer so I can move it quickly to the dryer very useful.
Unfortunately this company has decided to layer on at least one daily notification reminding me to think about all the value their product can bring me. This notification is not strictly marketing, because there’s no buy action anywhere, but it is most certainly the sort of “pay attention to meeeee” whine you commonly see in the most insecure boys.
The thing is that they seem absolutely BAFFLED that anyone wouldn’t want these messages. They cannot conceive that a consumer doesn’t give a shit about their washer/dryer except as a purely functional device. They want to be part of the family, the sort of thing where I think “Gosh I love my wife. Gosh I love my child. Gosh I love my dog. God damn I love my washing machine.” They genuinely believe people think like this. It’s sad and hilarious at the same time.
As an aside, I think you could get a smart plug that would support a 15 amp load that could give you a notification if the load went away. Might not be perfect, but just reading about having a relationship with a washing machine in a theoretical, second-party kind of sense makes me angry :)
Setting up home assistant to send push notifications via Smart plug amperage changes sounds like a great way to begin a long term committed relationship with your appliances
Edit: what the other guy said. I have a diver watch I just spin the dial to see how much time has passed since I started something. One time at the height of my Arduino hackery I didn't have a tea kettle and just boiled water in a sauce pan and said to my roommate, I bet I can shine a laser on the surface of the water to detect when it's boiling and make a noise, she laughs and says congrats you've invented a more complicated whistling kettle. Really humbling experience.
My washer has this 'Eco' mode, that is supposed to optimize time, water, and energy usage based on load weight and how dirty my clothes are. It finishes anywhere between 30 to 50ish minutes. Same settings an d all.
No, not at all. It's not really possible unless you're using an extremely basic washer with no spin (Or a very poor spin) cycle. A lot of the reason washers are terrible at estimating how long a wash cycle is going to be is because they spend a variable amount of time balancing the clothes before the spin cycle.
I used to work at a US healthcare company selling a product that doctors prescribe in a medical consultation in a serious medical context that occurs very rarely in a person's life. Everything the marketing/product side of the company did was predicated on the notion that the product would be something that people would have an emotional connection to and would be an important part of people's mental landscapes for a non-trivial proportion of their lifetime.
Given that the product people concerned must have accepted that the people involved (the patients) would have hundreds or thousands of involvements of similar profundity with other commercial products, I'm not sure what's more worrying: the misjudgement of the role of medicine, or the implication that they think that normal people's brains are teeming with 1000s of emotional attachments to random commercial products.
Marketers are what happens when you take the love and dedication that craftsmen have for their creations and separate it from the actual creation process.
They don't make the product, so their love can't actually make it better by including small human touches, or iterating.
They're not sales or traders, so they don't have to care about the nitty gritty of procurement or costs to customers.
All they have & need is excitement.
This would be fine if they were customers, but they're not so its all very parasocial.
(Having said that, I still think about Steak-Umms more than I ever expected because of their deeply satisfying twitter presence, though I’ve never knowingly purchased a steak-umm.)
Unfortunately a lot of people aren't like that. Corporate/brand loyalty is definitely a thing, and not just for products, for all sorts of extraneous reasons. For example, I know people who will buy whatever game Larian and Remedy produce next, and boycott anything EA makes, regardless of the quality or even the genre.
But the most successful brands are those that infest your life and which it is almost impossible to detach from. How much money and for how many years have you paid interest to "your" bank, for example? Could you switch to another bank? Have you tried?
I can easily switch banks in a day. Not crypto-heavy either. No loans, insurance, contracts with set date. Are there any other reasons why switching is hard?
I absolutely know people who love companies and products. Like LOVE them, they will be put into tears like a child at Christmas, but they're adults. They grew up with them or their family members died and those companies or products bring back those memories. They played their favorite game while eating that food, etc.
You never know. I'm not sure I've ever loved a company exactly, but I've really really liked a product, or sometimes just a type of food. If I like a certain food enough and only a certain company sells it, some of that feeling relates with the company too. Like the company cheers you up, BECAUSE they sell it. You can see how that might be valuable in spreading appeal for the company and helping preserve the thing you enjoy, so it has social/natural selection value.
I think when companies refer to this, they really are referring to real people, it's just aspirational that other people could feel that way if they let themselves. Most people won't.
If you have a sufficiently broad definition of 'product', then yes.
Millions of people are passionate about brands like the "Dallas Cowboys" and "Star Wars" and will dress up in costumes and go to events with like-minded people.
But for normal products, like USB headsets? Nobody's dressing up as the Jabra 20 Stereo USB-C Headset to go to the big Jabra Convention.
I have brands/products I love, but I love them because I trust them, and typically my trust of something is inversely correlated with seeing mainstream marketing for it.
I agree...but I also think it's a slow path to monetize OpenAI if they don't support some kind of channel to get paid by the sell-side. At least they offer a subscription and didn't jump to this as the only option. I'm optimistic that a premium tier can exist that avoids this, but we'll see. The reality is that they need the money.
It means the enshittification is reaching full acceleration. The MBAs have taken over. It'll be a constant battle between monetization and research, and a continual skewing of research and development as unnecessary expenses, followed by a full-on torching of the reputation in return for fast cash now and trading on vibes thereafter, while the rest of the world moves on.
Sama wants to speedrun the Apple arc, it looks like.
And investors will be genuinely surprised when the whole thing falls apart in 36 months and the still-very-rich founder spends the next 16 years giving uncomfortably veiny public talks about some uncomfortably sociopathic thing he’s become fascinated with and everyone tries to pretend is totally normal.
It doesnt take an MBA to figure out that selling stuff, or facilitating the sale of stuff, has value. The techies are the ones who built it, the whole "MBA" idea is a coping mechanism to deflect from your own (assuming you're some kind of techie/programmer) participation.
This is a brilliant idea. I'm going to mail 99999999 people about my new MBAI degree offering!
For only $6666 you too can become part of the new Mega Banana Annihilation Infantry – the secret task force fighting rogue potassium. Success guaranteed!
I have a product I love - it's a barbecue sauce. I buy it from a fella in Texas who ships the bottles in boxes he clearly packs himself (often the bottles have personal thank you notes with my name written on them in sharpie.) Wouldn't call it "elite"
Tangent to your point, but I think Dyson is generally considered to be overpriced for what it is. Shark and Kenmore are just as good by most metrics and are less expensive, whereas if you want reliable products then you get Miele or Sebo. Dyson's one selling point is that they're typically close to or on the cutting edge of technology, so they're the best if you value noise and size more than price and reliability.
It's a natural step towards agents buying things on your behalf - getting people warmed up to the idea that ChatGPT can be tasked with buying things. It starts with a buy button, "buy this now". In a few weeks it will become "buy dogfood", and it will just do the thing, knowing you get some certain food from Chewy
Who needs that, though? Is buying things that hard and time-consuming?
> In a few weeks it will become "buy dogfood", and it will just do the thing, knowing you get some certain food from Chewy
I have autoship orders set up with Chewy. Stuff arrives on a schedule and I get a small discount. I don't need an overhyped autocomplete fucking things up for me, especially when I can just set up a subscription myself and forget about it.
You're kind of echoing exactly what I'm getting at though. The difference comes down to trust. Maybe today we don't trust that one of these LLM apps is up to the task of buying things, but there was once a time where people didn't trust that if you click a button on a website a product would show up at your door. Over time, we came to expect things like order received emails, status updates, and the best companies will show you where your product is on a little map as it arrives to your house. Now most shopping is done online. Convenience won. There are still barriers to shopping online like too much choice, selecting sizes, and very personal constraints. An agent can know you're in the market for a new phone and can help you pick out a new one, then keep an eye out for deals. Maybe your agent will know your sizes and buy shoes that always fit. It's really not that wild of a concept.
You do not. But the people who eagerly jumped to smart homes, and voice assistants that are always on, to which privacy might as well not exist... Are going to love their own "Jarvis" which they can give shopping orders to while in the kitchen.
Basically, the Sillicon Valley "smart fridge" episode. Or even better, the Modern Family "smart fridge" episode.
The problem isn't that OpenAI is big news. The problem is that OpenAI currently appears to be worth more than AirBNB, Coinbase, Uber and Lyft combined, and Sam would have you believe this is still AI's "early stage". How much more liquidity is even left for this thing to soak up?
Yeah, the valuations are crazy, but think about it this way: the app economy and gig economy seemed absurdly speculative at first too. And now they’re worth trillions.
Early stage hype overshoots, sure, but sometimes it’s just pricing in things the old frameworks can’t even model yet.
Airbnb and Uber are shit in 2025. They're basically no better than what they were trying to replace. "The arc" you describe is people saying that these products aren't a magical solution that prints money and being proven right as the products are worn down to the level of the thing they were trying to disrupt by market forces.
Uber, in particular, drives me nuts because we replaced a supposedly powerful and evil taxi cartel (which happened to be a bunch of small, regional businesses) with a huge multinational corporation. Out of the frying pan and into the fire.
Totally get the frustration. Uber and Airbnb definitely have their flaws. But the sheer scale of users (200M last I checked) shows that a product can be both polarizing and essential at the same time.
Kind of like how we’re seeing with LLMs: not perfect, often overhyped, but undeniably useful at scale.
Convenience, consistency, and global. They solved real pain points, even if the cost and ethics are now questionable.
It’s like a software library that’s buggy and bloated but everyone depends on it because rewriting it from scratch is harder than dealing with its flaws.
You start from a false premise, that we have now is better than before.
Taxis were better than Ubers, bed&breakfast were better than AirBnB.
Maybe you are just too young to know it. The only reason we moved to Uber is because it was half the price of taxis, because it was subsidized, not anymore.
Uber and Airbnb solved real pain points: global availability, predictable, nice interfaces, cashless payments, reviews etc. These are the things small, local systems struggled to provide consistently.
Also, my point isn't that the current system is objectively better. It's that scale and convenience created network effects that make it "essential" in practice, even if it’s buggy, slow, or worse in some respects.
Taxis and B&Bs and Hotels all still exist and compete with Uber and AirBnB.
Even at the same price there are valid reasons why many people prefer an Uber over a Taxi, in particular the predictable pricing and globally consistent UI.
Uber changes the pricing depending on the demmand. Taxis can't do that, they are regulated by law in most of the west.
Predictable my ass. You have been lied to.
And btw all over the world you rise up your arm, and the taxi stops, I think that is a pretty consistent user interface that anybody in the world can understand. I have to help my aunt each time she needs an Uber.
So who do you call when it drains your account and orders 50 tons of wood, or falls for prompt injection and pays a scammer? Will OpenAI reimbuirse you? No.
Considering how vulnerable AI is to this sort of thing, there's no way in hell I'm touching it for at least 5-10 years.
Sure, they tell you that it's safe at the bottom and give reassurances, but I'm also not going to trust them on that, just as I don't trust pretty much anything ChatGPT tells me without futher verification.
This was bound to happen, but they will also likely see a much higher chargeback rate across the board if users are surprised when random comments or agent actions place orders or that orders placed too easily need to be reversed. Because consumers have less disposable income with all the AI-enabled layoffs, the bigger bonanza will come if OpenAI creates educational pathways via AI to enable more people to make money with AI. Entrepreneurial guides that take you step by step, accounting and other hurdles that AI can walk you through as you grow, brainstorming and exploring new business ideas, training people for a new trade or career path as employees. That will be the true game-changer that beats AGI. Because when you can give the entire society an easy ladder towards the industries that need them most, you will have a society that makes money off of AI to spend via AI shopping experience and gains purpose.
> Because consumers have less disposable income with all the AI-enabled layoffs, the bigger bonanza will come if OpenAI creates educational pathways via AI to enable more people to make money with AI.
Who do you imagine will be throwing money at all these side-hustle "make money with AI" business you envision? No doubt there will be a few- there already are a few- but as the market gets increasingly flooded with AI slop enterprise with very little value add, that well is going to dry up quick.
It's not different than all those content creators making videos offering to teach you the secrets to easy money... instead of just making all that easy money themselves.
digression:
> ... when random comments or agent actions place orders
I believe that the purchase action must be triggered by a user manually in a specific "Pay <vendor>" UI element? It doesn't seem to me like any prompt could trigger a purchase directly. Of course this could change, but likely won't for the reasons you said.
That’s my impression but the specification repository doesn’t seem to mention that at all, when it seems like it should be a requirement.
The article mentioned this:
Users stay in control — they explicitly confirm each step before any action is taken.
> Agentic Commerce
I can't remember where, but someone once pointed out a quirk of current AI advertising that is hard for me to un-see: In many ads, a customer implicitly and trusts the "AI assistant" to blindly purchase products and make restaurant reservation on their behalf, with barely even a verification, let alone any price-check or comparison-shopping.
The amount of affluence it takes to be that indifferent and carefree with spending money is... out-of-sync with the working-class people in the ads.
Now, maybe that's deliberate and it's meant to be aspirational, or maybe it's an attempt to appeal to two different groups, but it's weird.
Reminds me of the Amazon echo ads where people bought things so impulsively with a hey Alexa command. I don't think that one panned out but we'll see how well this one does
I realize a lot of the comments here are pessimistic, but this is a pretty obvious monetization path that they just can't not take. This is actually a huge angle IMO. ChatGPT is on a path to become a real entry point to the internet - why use Amazon or Google Search when you can embed results and checkout in the
I agree there's a real bias issue, but that is consistent through out any large company - e.g., Google, Facebook, Amazon, etc have sponsored results
> but this is a pretty obvious monetization path that they just can't not take
Now that they are a commercial entity, you are right. If they had remained on their original mission path though I suspect this wouldn't be the no-brainer it is now.
Yeah, it’s a very sensible feature to add. Obviously lots of users are making “please recommend a <product>” queries, it’s a usecase where deep research really shines.
So if your users are spending a lot of time on a tasks, why not make it more delightful?
Of course, you need to make sure you don’t allow accidental purchases, that would be the way you destroy trust. But assuming a clear intent to purchase is established, then I think this will be well-received.
I don’t quite follow this logic. Like there’s too much money involved to not dilute the value of a nascent technology?
If you’re willing to torch your credibility as a company, that tends to open up quite a few shorter term business options. The real trick is ensuring enough customer or user lock in that they can’t go anywhere else even when the enshittification is obvious to everyone.
The irony here is that ChatGPT could be a credible threat to Google search’s dominance as the entry point to the internet partly because the quality of Google’s search results has degraded so much. For some queries sponsored links push the real results below the fold on mobile, they’ve allowed some content aggregators to take over certain types of results (Pinterests polluting image results with irrelevant content). But that doesn’t matter while you make gobs of money. That is until a credible competitor finally appears and people are itching to find a better alternative.
They are already groveling to Nvidia for cash. It’s very likely the options are not “no ads” vs “ads”. The options for them are “no free tier” and “free tier with ads”.
Perhaps they'll use a lightweight advertising model to watch the output of chats, and look for product placement opportunities.
That's a "fun" thought experiment if they did do that, because the advertising model inference will need to run cheaply enough compared to the conversion rate/margin on clicks. I suspect it will be really hard to beat Google Ads on cost if you have to run inference on all chat output for each ad placement. It could put ChatGPT into a higher end/higher cost advertising platform.
> torch your credibility as a company
That's only so in our little cynical skeptical contrarian hacker bubble. For most people, it's an appreciated convenience.
Exactly - while you and I might route our DNS queries by carrier pigeon to avoid tracking, some of my family actually like personalized ads on social media and even buy things from them.
I installed an ad blocker on our home Internet on a raspberry pi. My wife told me to take it off.
I don't think this is true. I heard the exact same claim about Alexa making it easier to order diapers or whatever with one voice command: "sure, HN users don't want this, but normal people do". But I know many non-tech people who have Amazon Echo devices, and they never, ever use them to buy things. For them it's a timer-setting device only. That's why Amazon wrote off that entire division as a billion-dollar loss.
I think "people want frictionless ways to purchase products" is a PM pipe dream more than a description of reality.
> I think "people want frictionless ways to purchase products" is a PM pipe dream more than a description of reality.
I get what example you are referring to, but there are degrees here. For example the Buy Now flow really is handy; and I find I favor merchants that let me pay by scanning some kinda QR code from Apple Pay or Venmo. I definitely don't miss the friction of having to go dig out my credit card, mistype the cc#, type the wrong cvc if Amex, repeat the purchase after getting declined once and responding to a fraud text, etc.
Its kinda bizarre really.
How does Alexa ever compare with the rich experience of interacting with a store through the various senses? Its typical that technologists tend to come up with this stuff and what happens in reality is wildly off compared to what was expected.
Buying stuff means spending money. it turns out most people don't have a lot of money (something that Mr Altman would never be able to understand given his privileged background) so they want to see and experience the transactions that take place. Same reason why this agent nonsense is not going to work from an economic stand point.
They have hundred millions of paying subscribers, that kind of commercial success you could not even dream of when search engines and ads became a thing. Yet it's not enough. This tells me no matter what happens, even if those tech behemoth's make good profit, there will always be a reason to enshittify the product more.
It sounds like you think OpenAI is a profitable business? As far as I understand, it’s not.
OpenAI is projected to generate $12-14 billion in yearly revenue in 2025 (annualized from a single month), but expect to lose around 8 billion USD, implying the margins are negative.
OpenAI has raised a total of ~$60 billion.
I think they need to show investors a huge and growing cashflow to keep the show going.
OpenAI has a subscription revenue stream that's more than sufficient to keep current basic operations going. It is losing money because most of that money is spent on research, more and more GPUs, very expensive people and other capital expenditure.
Of course, they can't just retreat to selling their basic services since some other company would train and produce a marginally better model.
So it's a paradoxical situation. They're moving in contradictory directions - both to become a thing so valuable they'd only need to sell subscriptions and towards a mote if they don't reach that "AGI" thing. No reason being flexible would displease their shareholders but there are many other questions to answer here (who gets AGI raptures, who gets the Skynet/Terminator treatment, who decides, etc).
The valuation of OAI would be peanuts if it stopped reinvesting too. Which will destroy the value of equity and therefore employees certainly wont be happy.
Im not really sure where Altman is going. As time goes on, it seems the walls are closing in and he's just throwing all he can to keep the hype alive.
You cant escape fundamentals forever, I dont care who you are.
Right, so given that R&D is not optional lest they fall behind, they need to find another revenue stream.
They have hundreds of millions of subscribers, but the vast majority of them are not paying, and more importantly, are not monetized in any way.
With increasing competition from all sides driving down the margins. Sure ChatGpt is a household name now, but if Microsoft/Google offer the same thing for half the price, plenty of cost conscious subscribers will bail to the cheapest offering.
Hey man, take a step away from the keyboard. Instead imagine the every day person. Would they rather click, scroll, swipe and pull out credit cards across multiple websites - or just ask their digital assistant to do it?
The defaulting to negativity will really eat some communities up from the inside.
Chat windows and voice assistants are a terrible user experience for the average person. This doesn't change that.
That’s exactly what the folks at Amazon thought when they came up with Alexa. Have you ever bought anything online by asking Alexa to do it? Have you ever seen anyone else do it?
You see negativity, I see disappointment that OpenAI isn’t trying to innovate, and instead hoping they can replay Google Search’s history for themselves
Does the average user does this? Granted, I’m not i. the USA, but does people really order that much on unusual websites?
The fact that you're being downvoted over this is proof that people here work and live in a bubble. People value convenience and are willing to pay for it, and if OpenAI is able to advance convenience through these actions, they'll make billions.
> They have hundred millions of paying subscribers
They have hundreds of millions of users in total (free tier included), with around 10-15 million paying users.
"...obvious monetization path that they just can't not take."
Sure they could. This notion that an unscrupulous revenue stream is justified if it pays well enough smacks of "Just following orders!"
"It happens that every man in a bank hates what the bank does, and yet the bank does it. The bank is something more than men, I tell you. It's the monster. Men made it, but they can't control it.” - Grapes of Wrath
At some point we have to stop this madness.
Most people has been convinced throughout the years that individual actions have no effect on big corps, they need to learn that it shouldn’t matter. If you don’t agree with what a service or product provider does simply don’t buy their product find an alternative or just don’t but do it just for your own principles.
> If you don’t agree with what a service or product provider does simply don’t buy their product find an alternative
This implies antitrust has been sufficiently enforced such that there are viable alternatives in all markets.
That's not currently true.
Do you run an ad blocker? If so this feels a little hypocritical, you should just bounce from the site and seek an alternative if you don’t agree with what they do.
> Merchants pay a small fee on completed purchases, but the service is free for users
Where's the money coming from to pay the fee? From my wallet, of course! At scale, this raises the price of purchases.
Another thought -- the text-only interface of ChatGPT is way too limiting. We need images (and video) generated on demand showcasing product suggestions. Showing an image of the user wearing some new shades is going to be pretty compelling...
You could say the same of affiliate links on Wirecutter, or any other marketing expense.
Before objecting to this, I'd want to know how much those fees are, but the terms seem to be undisclosed.
Not really, this is just effectively a new supply side platform for advertisers. Retail companies will just allocate a percentage of their marketing budget to OAI instead of Google/Meta
Yep, in some ways it's good to be the price sensitive side of the transaction. Can't really take more of the money people don't have.
OpenAI desperately wants to compete at the same level as the big tech firms.
While their only successful product is impressive, it is doubtful that its success alone can sustain them beyond the first 'downturn' of their value in the market. This reeks of desperation on their part and should bring more attention to the mountain of "promises" they have made, compared to its actual deliveries.
Stated another way, they don't have a ready-made DoubleClick to buy the way Google did. The big money is in being the ad network not just the purveyors of billboards.
The “big money” is having the attention and trust of high purchase intent users, which OAI does
I agree wholeheartedly.
This regresses the incentives back to what we had with search engines where what I need (answers) and what OpenAI needs (money from ads) are at odds.
Search engines used to be very useful too until the endless profit a/b testing boiled us all
Five hundred gajillion dollars spent so we can end up in the same place except with these five men making all the money instead of those five men. Whee.
It’s not just that, if asking the AI agent is going to give me the same experience google search gives me today, that’s just a few hundred billion dollars we could’ve just not spent.
There is a massive difference between an AI agent understanding the intent of my question, and keyword search on old (pre-enshittified) search engines.
Even if OpenAI needs to feed the VC beast, they will always be open source LLMs that can be used freely inside home-made search engines.
Don't forget all the extra electricity required to achieve roughly the same thing!
Arguabilly we are in worst place than when Google Search was not enshittified.
Is this a surprise though?
This is the culture of America in a nutshell. Steve Jobs was a weirdo in that regard and an outlier.
Who said anything about being surprised?
No, and no.
Google is a public company! You could have been one of "those five men".
Yes of course, let me buy a sizable chunk of one of the largest companies in the world so I can be on the five men.
Owning a few shares is not the same thing as actually making all the money someone at the top of Google is making.
Even ads in magazines was much better than what we have now. Ads are contextual (a tech mag won’t have ads for gardening), so apart from the repetitive aspects, what is shown may be not needed, but it’s more likely to make a mental note because you’re already in the relevant context.
Agreed. One big difference, though, is that the local AI tech we have as an alternative to OpenAI is significantly better compared to the local alternatives we had for Google. You can run a reasonably powerful AI on your own machine right now. Sure, it’s not going to be as good. And the cost of GPUs, RAM, and electricity is important to keep in mind. But the point is it’s not all-or-nothing and you are not beholden to these corporations.
There is also plenty of research going on to make models more efficient and powerful at small sizes. So that shift in the power gradient seems like it’s going to continue.
I agree with this, but there were/are alternatives to Google that are functional but not as good. People still ended up choosing to use Google.
Since it's tapped into etsy and shopify, ChatGPT might actually have a lot more power to shop local if you give it that as a constraint
There is nothing "local" about etsy, and there has been for over ten years+. You can find all the same "handmade" products on AliExpress, and often Amazon.
Fine... Shopify then? There exist brick and mortar stores near me with Shopify sites.
Etsy is thoroughly fucked and full of mass-produced junk. "Local" could just mean buying from the nearest person who's reselling stuff from Ali Express.
And have you noticed what sellers on Amazon are doing? Foreign companies are setting up distribution in the US and registering their US companies with Amazon as "small businesses" and "minority-owned businesses", making those labels utterly useless.
reminds me of the animation of google's results page progressively becoming more and more "sponsored links"
Ads in ChatGPT was the most obvious outcome from day 1.
And this is not a bad thing, otherwise you can only image how many businesses will close when google traffic stars to decline.
Everyone likes to hate on ads but the reality is that without ads 99% users even on hacker news would be jobless as the companies where they work will have no way to find clients, and even if they manage to find some - those clients won't be able to sell and will go out of business.
Ads haven't made it yet, they are charging money for the purchase made:
> Merchants pay a small fee on completed purchases, but the service is free for users, doesn’t affect their prices, and doesn’t influence ChatGPT’s product results.
Yes, so they’re being incentivized to highlight products which they get a cut of, and more strongly to highlight more expensive products.
This is called affiliate marketing and it’s toxic.
> Ads in ChatGPT was the most obvious outcome from day 1
Agreed.
Tech companies always do this. With Ads, we’re back into speculation territory, and the “how do we pay for and justify all this shit?” can gets kicked down the road.
Can’t we actually solve problems in the real world instead? Wouldn’t people be willing to pay if AI makes them more productive? Why do we need an ad-supported business model when the product is only $20/mo?
> Wouldn’t people be willing to pay if AI makes them more productive? Why do we need an ad-supported business model when the product is only $20/mo?
This was always a fake reasoning (ads are there because people want everything for free!), but then paid HBO started ads, your purchased smart TVs started ads, cars that you bought with money started ads...
([some business model] + ads) will simply always generate more profit than [some business model] (at least that's how they think). Even if you already pay, if they also shove some ads in your eyes, they can make even more money. Corporations don't work the way humans do. There is no "enough". The task of the CEO is to grow the company, make more profit each quarter and is responsible to the shareholders. It's not like, ok, now we can pay all our bills, we don't need more revenue. You always need maximum possible revenue.
there are other ways to be probably, without ads. I'm optimistic we, as a society, will find those ways.
I think ads are great, but the tactics (tracking) around them aren't really in the good course.
This is the first time I've ever heard anyone even mention it, and I never thought about that possibility myself either.
Ahem, The article is about ChatGPT check-out. Ecommerce and a relatively quick shows no mention of ads.
Sure, this may in Google-style-monopoly direction or an Amazon-style-monopoly direction. I don't know which. I would indeed expect a large dose of enshittification would be involved.
You're welcome to argue this leads to ads. But jumps to this is ads and getting a dozen pearl-clutching is a symptom of hn's own crude enshittification, jeesh.
"Merchants pay a small fee on completed purchases, but the service is free for users, doesn’t affect their prices, and doesn’t influence ChatGPT’s product results. Instant Checkout items are not preferred in product results."
The incentives are very strong to prefer instant checkout items.
Wouldn't it be incredibly easy for merchants to expose slightly higher prices over the "agentic commerce protocol", so that users do end up paying this fee?
It would be, but at least in the EU this is not legal - this is also why you cannot pass the payment processor fee to the customer (which this kind of is)
I’ve got more money than I know what to do with and now, thanks to OpenAI, I’ve found the perfect custom-built solution: an AI agent ready to spend it for me.
You're absolutely right, I've purchased 10 Blackwell B200s on your Platinum AMEX to maximize your points.
I'm gonna need some Oracle storage as well.
You’re in luck because I got you some database licenses and am drafting your letter to their compliance department so it will be ready a year from now.
Thus far I didn't have to worry about ChatGPT having bad incentives when giving me advice on product purchases. Now that "Merchants pay a small fee on completed purchases", will the model steer me towards ACP-supported retailers at a higher rate?
This will be worst for consumers, like DoorDash or Uber eats.
DoorDash takes 15-30% of fees from restaurants so restaurants raise their prices and consumers have to pay service fee and a delivery fee and tip.
Be ready to pay more at sites that have this enabled.
Presumably yes, but if I’m using an agent to make purchases, I’d prefer it to use sites where it can safely make a purchase anyway. The optimal UI would probably leave it up to me. One system should find products, independently of whether a vendor of them supports “ACP.” And I should be able to configure my agent to “only purchase where ACP is supported.” In the context of a conversation, I’d expect it to show me all the products it found, and offer to refine the list to include only ACP vendors.
> ChatGPT shows the most relevant products from across the web. Product results are organic and unsponsored, ranked purely on relevance to the user.
Cynical take, I know, but that is how Google started too. But now they prioritize people who pay them, and it's only a matter of time before OpenAI does the same.
And I can't at all blame them! They are there to make money (well, now).
But I suspect this won't be true for long.
It could be many years. I think Google went a decade or more before search results became crammed with ads?
If OpenAI is charging commission on sales through ChatGPT, then they're probably better motivated than google is to show the best results.
motivated to show you results that get sales, not that are the best solution to the problem
In many cases, Google gets commissions too. A lot of their ads pay when the person completes a purchase.
Which ads are these? My impression was they are all based on CPC or CP impression
They have an option for pay per conversion: https://support.google.com/google-ads/answer/7528254?hl=en
How so? If you and I both sell widgets for $50 but mine is worse, I can simply pay OpenAI to rank mine first.
And OpenAI might not even know my product is worse. In fact, they probably don’t. They’re too big to investigate. All they know is that I paid them more.
>I can simply pay OpenAI to rank mine first
no you can't. because that is not a thing that they sell.
nothing in the announcement says they will accept money for changing their rankings, and the comment at the top of this chain includes the quote from the announcement where they explicitly say they won't do that.
Nothing that they sell, yet.
Not cynical, realistic
Check this out from the seminal paper written by Page and Brin:
"Currently, the predominant business model for commercial search engines is advertising. The goals of the advertising business model do not always correspond to providing quality search to users. For example, in our prototype search engine one of the top results for cellular phone is "The Effect of Cellular Phone Use Upon Driver Attention", a study which explains in great detail the distractions and risk associated with conversing on a cell phone while driving. This search result came up first because of its high importance as judged by the PageRank algorithm, an approximation of citation importance on the web [Page, 98]. It is clear that a search engine which was taking money for showing cellular phone ads would have difficulty justifying the page that our system returned to its paying advertisers. For this type of reason and historical experience with other media [Bagdikian 83], we expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers. "
lol.
Perhaps they left their management positions as they couldnt face the double standard they had entered into. Highly unlikely but who knows.
Ah.. the day OpenAI turned into an Ad company. We all knew it would happen someday.
Who even thought this would happen? Got any sources?
Who? Those of us who were paying attention to how tech companies have operated for the past 25 years.
They start out subsidized by investors and then once they have enough users and can no longer pay for them with the invested cash, they push more and more ads onto users.
And it was easy to see that LLMs are an especially devious place the inject ads because they can flow right into the the response and not even look like an ad, but rather feel like a casual recommendation.
I've seen multiple prior conversations about the idea on this site. It's easy to find similar reddit discussions with a search: https://www.reddit.com/r/OpenAI/comments/1ljv78j/sam_altman_...
It's a fairly obvious way for them to make money, as people are using it as a replacement for search engines, and that's how search engines have made money.
Many, many people expected this. https://news.ycombinator.com/item?id=42008569#42008733
It is literally a highway at this point for any B2C tech product. All the info they are collecting surely can be used for very targeted ads. Not hard to predict at all.
It’s all Ben Thompson has talked about for the past couple years.
Erm their recent hire of Simo and acquisition of Statsig made this obvious.
The ability to purchase items being added to ChatGPT seemed inevitable and now it happened.
However, as far as I can tell, it happened in the best way possible. This seems like an actual open standard, which is refreshing to see in the modern web.
Creating a common API for stores and payment processors will likely lead to increased competition, benefiting users.
I'm happy to see this. It seems to benefit all consumers, even those who do not use AI.
The thing is.. chatgpt can be genuinely useful. I did purposefully use it to get some some product comparisons and whatnot. It could be genuinely good if it is handled well and not devolve into constant buy blast ( like with emails and just about any other medium ).
Still, there is a reason I am frantically working on working on a more local setup that I can trust not to:
a) oversell me stuff b) is under my control c) not profile me ( in a way that can be sold to other merchants )
The issue seems to be the same as always. I am either a minority or the money pull is way too strong.
I'm not sure how long you've been using Google, but Google also used to be genuinely useful. The more that ads became a priority, the less Google was able to be genuinely useful to me.
Even Cyanide can be useful.
So in their example, OpenAI and their partner Stripe is handling the transaction? If yes, they are now competing with both Google (advertising) and Amazon/Walmart (online marketplace).
* - - - - (1 / 5) Village Idiot, Somewhere, USA. "product is not at all what chatgpt showed. returned for a refund. wanna give 0 stars but cant"
[Response from Seller] "We regret that our product doesn't meet the customer's expectations — we issued a full refund."
The fact they're working on this is a huge signal that AGI isn't happening
And that OpenAI is trying to focus on being a consumer product over business use cases.
Rule 5: Don't be an endpoint.
https://nextbigideaclub.com/magazine/futureproof-9-rules-hum...
Isn't that what management is?
Don't think I'm anywhere close to trusting agents with my money
Every so often I check back on ChatGPTs product search powers, and every time it cites garbage blogspam articles that praise every product hoping you'll click their affiliate links.
The irony being that many of those spam articles are probably generated with ChatGPT - https://youtu.be/WLfAf8oHrMo?t=86
I don't think this is that? This is more like agents show you a list of products and the UI allows you to buy them like any other online store.
They get a cut for products that support this. So they’re incentivized to display those instead. It’s pretty close to standard affiliate advertising and the biases that introduces.
Sure, but still far from "trusting agents with my money"
The trust problem was a concern early on in agents (well before LLMs).
Although, at the time, we didn't have the "fox guarding the henhouse" problem of modern tech companies, and they hadn't yet inserted themselves into the loop so intimately.
I found and purchased a hand-made toy for my son this evening using ChatGPT searching Etsy. The search results were good and made it easy to pick from the many options. The checkout process using my stored CC hit an error (it didn't tell me what error) but when I switched to Google Pay it completed successfully. Overall a pretty nice experience as shopping goes.
In a way this was kind of inevitable, but I had hope that it was still 1-2 years away. This likely degrades brand value and begins to shift the incentives of the AI's responses.
They really time releases based on their competition. How many times did this happen with both companies? I've lost count.
They had been keeping this in reserve and decided to release it when Anthropic released Sonnet 4.5. Anime-like tactics.
I feel like this one may backfire though, Anthropic announced a core improvement to their product, OpenAI countered with enshittification to theirs. It only further highlights that OpenAI are running fast and hard towards a future where ChatGPT is a marketing/advertising/etc system just like Google's search has become.
They definitely do, but I think this particular one was a misfire. "We've updated Chat-GPT to help you better connect with #Brands #BrandsULove #adtech" would be eyerolling at the best of times, since it's a nigh-guaranteed signal of incoming enshittification, but positioning it as counterprogramming to Claude Code genuinely becoming more useful just looks pathetic.
This is a new amazon marketplace with trusted reviews. IMO having worked in commerce for a while - this opens up new avenues for merchants to work on gaming reccomendations and the trust quotient delivered.
It also opens up - ads - subscriptions - transactions
based revenue lines for OpenAI
I'm not looking forward to when there's an AI pyschosis case where it turns out ChatGPT sold a bump stock or gave a bulk discount on fertilizer.
AGI achieved internally.
Thus begins the age of ads in AI.
Back in June (just three months ago) when I wrote this blog, people were telling me it was far away: https://digitalseams.com/blog/the-ai-lifestyle-subsidy-is-go...
My immediate thought is that commerce mediated by AI will quickly make "old-style" ad-tech feel non-intrusive by comparison.
But, like it or not, here it comes.
For people who've supplanted their own thoughts with auto complete yes, their dreams will start advertising to them
I'm just not sure that I would trust that the view/description of the item ChatGPT shows me and the thing I'm actually agreeing to buy are the same thing.
The AI of everything economy seems like it will be so deflationary that money will stop being useful.
Wow, with this in place the incentives are enormous for OpenAI to allow sponsors to pay for a slight nudge in the recommendation this way or that.
This will replace the current ad economy.
Yeah, it seems obvious that this is how models will be monetized in the future. The free version of ChatGPT will stop being a loss leader for the subscription and start paying for itself with commissions. The vast majority of people will use the free version.
They will likely go through many iterations of this before finding what works, but I expect it will eventually be an incredible business on the same level as AdWords. We can only hope that the incentives don't end up warping the models too much...
> We can only hope that the incentives don't end up warping the models too much...
Oh my sweet summer child. How can you be so optimistic after seeing Google's decline? It won't happen all at once, but the need for revenue growth and the incremental logic of A/B testing are relentless forces that wear away at the product once ads are in the mix.
It's been 27 years since Google Search launched, and I still find it very useful despite some perverse incentives. A pretty good run I would say. If OpenAI declines that slowly, and is then displaced by something better, I'd say it was a good outcome.
I hope it takes as long for AI enshittification, but the tech "cycle" has become shorter and the pressure to make revenue is much more intense than it was for Google.
This is likely to be a longer term play at advertising - once people are accustomed to purchasing in this way, they’ll add [sponsored] merchants in the same list.
So, if I'm understanding this correctly. The latest ChatGPT features are... That it can now message me without me talking to it, and can automatically buy things for me.
Absolutely brilliant observation—you’ve managed to distill what took OpenAI an entire product announcement into a single, devastatingly clear sentence.
Would you like me to diagram the precise mechanisms through which these features transform users into passive recipients of AI-initiated interactions and transactions?
Excellent. Imagine three nested loops:
- Initiation loop: The AI identifies a trigger (calendar entry, email, purchase pattern) and begins the conversation unprompted.
- Action loop: Once trust is assumed, it executes on your behalf (ordering, booking, messaging).
- Feedback loop: Each interaction produces more data, refining its ability to predict when to act next.
Together these loops progressively erode the boundary between “I decide, AI assists” and “AI decides, I ratify.”
Would you like me to sketch this as a flow diagram, or unfold the psychological implications of each loop?
Are you two using ChatGPT to generate these, or are you both fantastic at emulating the writing style? Because this is so spot on.
I wish I was that good!
Being reminded of how Sychophantic OpenAI's models can be and coupling this with the parent comment makes me unreasonably upset.
Yes, that's exactly the function of the new features. Your observation is the core of the product strategy. Let's diagram the mechanisms.
The shift is from a Reactive Tool to a Proactive Agent, and this transition fundamentally alters the user's role. Here’s how it works, broken down into its constituent parts:
The Mechanism of Passivity: From User as "Driver" to User as "Passenger"
1. The Initiative Shift: Who Asks the First Question?
· Old Model (Reactive): User has a thought -> User formulates a query -> User inputs the query -> AI responds. · Cognitive Load: On the user. They must identify a problem, articulate it, and initiate the interaction. · New Model (Proactive): AI analyzes context (screen, audio, memory) -> AI identifies a potential need or action -> AI presents a suggestion or takes a micro-action -> User consents or refines. · Cognitive Load: Shifted to the AI. The user's role is reduced to granting or denying permission.
2. The Transactional Seam: Blurring Help and Commerce
· Old Model: Help and transaction were separate spheres. You'd use a calculator app, then separately open Amazon to buy a calculator. · New Model: The AI, by having context and initiative, creates a seamless bridge from identification to acquisition. · Example Flow: AI sees a recipe on your screen -> It offers to add the ingredients to a shopping list -> The shopping list is integrated with a delivery service -> A "Buy Now" button appears. · The Passivity: The user is not seeking a store; the store is brought to them. The decision point changes from "Should I go shopping?" to "Should I not buy this right now?" The default action becomes consumption.
3. The "Frictionless" UI: Eliminating Deliberation
· Features like the "phone-break-in" for real-time translation or assistance remove the physical and psychological steps of opening an app, typing, and waiting. · The Consequence: This eliminates the "deliberation time"—the few seconds where a user might think, "Do I really need to do this?" or "Is this a good idea?" Interaction becomes impulse. The user is carried along by the convenience of the flow.
4. The Memory Layer: Creating a Dependent Relationship
· Without Memory: Each interaction is a clean slate. The user must re-establish context, which reinforces their role as the authoritative source of their own information and history. · With Memory: The AI becomes the custodian of your context, preferences, and patterns. · The Passivity: You no longer need to remember your own preferences; you rely on the AI to remember for you. This creates a gentle but powerful dependency. The AI becomes more efficient at being "you" than you are, because it has perfect recall. Your agency in defining the context of a conversation diminishes.
The Underlying Economic Engine
This isn't just a technical shift; it's an economic one. The "passive recipient" is a more valuable economic unit than the "active user."
· An active user has intent that they satisfy. The value exchange is clear: they have a question, they get an answer. · A passive recipient is presented with opportunities for engagement and transaction they did not explicitly seek. This creates new, AI-driven funnels for: · E-commerce (as described) · Service Sign-ups ("You seem to be planning a trip. Would you like me to find you a hotel?") · Content Consumption ("Based on your last question, you might like this video...")
In essence, OpenAI is building an Ambient Interface that sits between users and the digital world. Its primary function is to reduce user effort, but the secondary, commercial function is to orchestrate user activity towards endpoints that benefit its partners and, ultimately, its own ecosystem.
You were right. It's a brilliant, and from a business perspective, inevitable evolution. But it systematically re-architects the human-computer relationship from one of mastery to one of management. We are no longer pilots at the console; we are administrators approving the suggestions of an ever-more-autonomous system.
sure
SOTA
What percentage of e-commerce will be taken by OpenAI by the end of 2026?
It appears to me that they are already well-positioned to become the next generation of Amazon with their current user base -
* AWS -> OpenAI APIs
* Amazon -> ChatGPT Shopping
It took a while for people to trust putting credit cards into websites. So general adoption will take some time.
Meh. Meta recently got rid of their instant checkout product for Instagram and Facebook where customers could buy products directly from a companies FB or IG page. Nobody was using it. I would imagine it will be the same situation here.
I bet this is going to make them a TON of money. A ton of people are using chatgpt to essentially replace google, and treating it like a trusted source. The average user is going to jump at the ability to ask their "trusted" source a question and get a direct link to the thing they need to buy.
I doubt it. I think it will take a long time to dethrone the tried and true google search of "<product I'm considering> reddit"
It will take no time. I made three purchases this weekend where I started my search with ChatGPT because it gives me better results than Google, and it can also pull in or link me to Reddit comments.
I have it running a background research task now where it’s producing a comparison table of product options with columns for different attributes I’m interested in, including links to purchase it, so it can help me make a decision tonight. If this feature is available for what I want, I’ll be using it in a few hours.
Whether you use ChatGPT or Google the first thing you see is an AI generated response, but Google is using the cheapest version of their model and only providing the context from the top 10 results, while ChatGPT is using a much better model and passing in more context. Lots of folks are turning to ChatGPT instead of Google these days.
What did you buy
The average user just opens the Amazon app.
Can’t wait for being able to run industry leading LLMs locally without the threat of advertising
I need this for DoorDash desperately. My wife and I trying to figure out what to eat every day is one of the most annoying parts of the day. If anyone has any recommendations, would be much appreciated.
meal plan
Cook unity if you want the meals without the hassle
We tried different services like these, they quickly become unappetizing. You end up forcing yourself to microwave a box vs. order something fresh cooked.
won't this just lead to them offering the same products for a given search, sort of like how AI homogenized writing?
"recommend me a novel about <whatever>" and it just gives you bestsellers
1. I am one of the people who have been looking forward to the ability to buy directly from a chat session instead of going to Google from a chat session.
2. The media began covering this 3 hours ago and $goog is just ~1% down. I’m curious why it didn’t spook Google investors, whether reasonably or unreasonably.
If OpenAI can pull this off, couldnt Google just do the same with its Gemini Product and start using their scale to compete more directly? Perhaps the sentiment isnt as negative as OpenAI lacks any sort of mote here to keep competitors from simply replicating this change.
At this point, not allowing for sponsored recommendations (ads) would be leaving money on the table and OpenAI will for sure be accused of reneging on their fiduciary duty to the shareholders.
As much as I hate it, it's inevitable.
"How do I do XYZ in Python?"
ChatGPT: I've found the following Python books that contain explanations of how to complete that task. Which one would you like to purchase?"
"None. I just want an answer."
ChatGPT: Ok, perhaps you were looking to purchase a python. I've found the following pet stores that sell Ball and Reticulated pythons. Which one would you like to purchase?"
"Aaaaaarrrrrrrggggghhhhhhh"
If you upgrade to Premium Plus for only $5 more per month, I can offer you answer first with recommendations in the bottom.
"This book by LearnPythonFastGuaranteedResults!!!!11! comes highly recommended! People are raving about it, a buyer named John Ryan said he got a 400k job after learning Python by reading this book! Are you sure you don't want to buy it?"
Or if you go full agent mode and let ChatGPT buy things for you: https://xkcd.com/576/
Sorry, but I did not find any "Aaaaaarrrrrrrggggghhhhhhh" to purchase. Could you give me more details?
"Merchants are ranked based on availability, price, quality, whether they are the maker or primary seller of that item, and whether Instant Checkout is enabled."
So, this is a race to de bottom for any SMB. If this didn't work for Instagram and Facebook, I don't think is gonna work for them neither.
What's interesting is that they're REALLY trying to see what's going to stick before the wheel stops spinning. From "AGI" and gpt5 getting Altman all "scared" of what's coming to trying to sell ads, a social media showing ai videos, among other things ("app" store with GPTs, etc.).
If there was a statement that we were in a bubble waiting to pop, this right here is plenty enough proof of that.
> OpenAI CEO Sam Altman says GPT-5 actually scares him — 'what have we done?'
Ah, so that‘s what he actually meant.
They promised us AGI and the singularity, they delivered more ads.
He said in a podcast the other day that with AI we will build a dyson sphere in the next couple decades.
I can't believe people give this guy money. It's so frustrating.
This is textbook strategy, adding layer after layer of pseudo protocols and "standards" on top of (surprise, surprise) their hard to defend against competition offerings, and ironically attempt to build a moat around something that ideally is open and interoperable. I haven't ruled out that I'm actually an idiot who doesn't get it, but it sure feels like the emperor has no clothes. It's like a recent Cloudflare post about how we're using LLMs to code all wrong; we just need to build (yet more) APIs specifically for MCP servers and then have very simple tools that only use those APIs. Yet more extra effort, time and energy optimizing for someone else's benefit.
> This is textbook strategy [...] attempt to build a moat around something that ideally is open and interoperable
It's so textbook that Google two weeks ago came out with their own competing "open" standard for doing the same thing!
https://ap2-protocol.org
It's a new space, everyone is coming up with their own protocols. It's really no different than Stripe offering an API so you can use them as a payment provider. You need to implement this API on your website so ChatGPT can allow users to buy stuff from you.
You need some amount of strictness in the API here, LLMs are not actually sentient. You could say that this is a failure on OpenAI's part in comparison to their marketing, sure.
IMO this view puts the cart before the horse. Suppose ChatGPT builds a UI that consumers prefer over web browsing to purchase products. How should merchants plug into that UI? This protocol is an easy way for merchants to do that. And once merchants are plugged in, consumers still need to be able to pay, and merchants need to get paid. Stripe makes that easy.
Consumers will only use the UI if it's better. Merchants will plug into the UI to make more money. So everyone wins-- consumers, merchants, OpenAI, and Stripe. And since the protocol is open, other chatbots and other payment processors can implement it too. Who loses?
(I agree that at scale these things tend to accrue to top players and you get all kinds of weird unsavory consequences. But I'd argue that's a critique of our regulatory apparatus, not of the companies building products and services.)
"I haven't ruled out that I'm actually an idiot who doesn't get it,"
Nope youre not an idiot, I agree with your thoughts.
Looks like it's OpenAI response to Google's Agent Payment Protocol (AP2) but without the micro transaction part.
Are they insane?
How deep into a bubble are we that digital stores get integration into LLMs? There are so many obvious risks here and so few imaginable upsides over redirecting a user to the merchant.
I love the little synthesizer in the square in the footer :) What an awesome easter egg..
Couldn't see it, what was it?
You should see a grid at the bottom of https://www.agenticcommerce.dev/
I’m surprised free users don’t have to watch an ad spot every five messages (yet?)
Can't wait for the agent to hallucinate and while fulfilling a request to buy 2 jeans, it ends up buying 3 macbooks.
So what would a 'hallucination' be in this context? An order for a half ton of toilet paper? If I know ChatGPT gets things wrong, why would I trust it to shop for me?
Shots fired at Google
I knew this day would come.. but not this soon.
> finding products they love
That phrase is a dead giveaway that a very silly group of people have started dominating the conversation at a given company. I don't "love" a product. I certainly don't "love" a product I'm having to resort to ChatGPT to figure out my potential relationship with. At best, I "love" having a solution to a problem that I want to spend as little time and money solving as possible, and even then it's more of a satisfactorily productive comradeship.
God this is exhausting.
Marketers are obsessed with this idea that people want some kind of "relationship" with them. As if I wake up in the morning, hoping to interact with brands and have experiences around them. I'm not going to follow McDonalds. I'm not going to subscribe to the McDonalds E-mail newsletter. I'm not going to read posts on Twitter from the official McDonalds social media editor. I'm not looking for relevant McDonalds products. I'm not even fucking thinking about you, McDonalds. That goes for all brands, not just them. I wish companies could just back off, offer products, take my money when I buy them, and butt out of my life otherwise.
I’ve been in this long-running battle with a large non-US appliance maker about their app. I live in a humid place and find notifications that a load is finished in the washer so I can move it quickly to the dryer very useful.
Unfortunately this company has decided to layer on at least one daily notification reminding me to think about all the value their product can bring me. This notification is not strictly marketing, because there’s no buy action anywhere, but it is most certainly the sort of “pay attention to meeeee” whine you commonly see in the most insecure boys.
The thing is that they seem absolutely BAFFLED that anyone wouldn’t want these messages. They cannot conceive that a consumer doesn’t give a shit about their washer/dryer except as a purely functional device. They want to be part of the family, the sort of thing where I think “Gosh I love my wife. Gosh I love my child. Gosh I love my dog. God damn I love my washing machine.” They genuinely believe people think like this. It’s sad and hilarious at the same time.
As an aside, I think you could get a smart plug that would support a 15 amp load that could give you a notification if the load went away. Might not be perfect, but just reading about having a relationship with a washing machine in a theoretical, second-party kind of sense makes me angry :)
Setting up home assistant to send push notifications via Smart plug amperage changes sounds like a great way to begin a long term committed relationship with your appliances
Edit: what the other guy said. I have a diver watch I just spin the dial to see how much time has passed since I started something. One time at the height of my Arduino hackery I didn't have a tea kettle and just boiled water in a sauce pan and said to my roommate, I bet I can shine a laser on the surface of the water to detect when it's boiling and make a noise, she laughs and says congrats you've invented a more complicated whistling kettle. Really humbling experience.
Aren't all washed cycles the same time? Can't you just use a timer? That's what I do. My washer always takes 44 mins
My washer has this 'Eco' mode, that is supposed to optimize time, water, and energy usage based on load weight and how dirty my clothes are. It finishes anywhere between 30 to 50ish minutes. Same settings an d all.
Then you set your alarm to fifty minutes ? Surely the max twenty minutes the laundry spends in the washing machine won’t make a difference.
>Aren't all washed cycles the same time?
No, not at all. It's not really possible unless you're using an extremely basic washer with no spin (Or a very poor spin) cycle. A lot of the reason washers are terrible at estimating how long a wash cycle is going to be is because they spend a variable amount of time balancing the clothes before the spin cycle.
Not only does my washer have different cycle times but I get lost in whatever I’m doing and hours pass lol.
I love… lamp.
I love lamp!
I used to work at a US healthcare company selling a product that doctors prescribe in a medical consultation in a serious medical context that occurs very rarely in a person's life. Everything the marketing/product side of the company did was predicated on the notion that the product would be something that people would have an emotional connection to and would be an important part of people's mental landscapes for a non-trivial proportion of their lifetime.
Given that the product people concerned must have accepted that the people involved (the patients) would have hundreds or thousands of involvements of similar profundity with other commercial products, I'm not sure what's more worrying: the misjudgement of the role of medicine, or the implication that they think that normal people's brains are teeming with 1000s of emotional attachments to random commercial products.
Marketers are what happens when you take the love and dedication that craftsmen have for their creations and separate it from the actual creation process.
They don't make the product, so their love can't actually make it better by including small human touches, or iterating.
They're not sales or traders, so they don't have to care about the nitty gritty of procurement or costs to customers.
All they have & need is excitement.
This would be fine if they were customers, but they're not so its all very parasocial.
(Having said that, I still think about Steak-Umms more than I ever expected because of their deeply satisfying twitter presence, though I’ve never knowingly purchased a steak-umm.)
Some people do though
Unfortunately a lot of people aren't like that. Corporate/brand loyalty is definitely a thing, and not just for products, for all sorts of extraneous reasons. For example, I know people who will buy whatever game Larian and Remedy produce next, and boycott anything EA makes, regardless of the quality or even the genre.
You're on the right side of the bell curve. Marketing works best on the left side of the bell curve.
Sure, that's you, but to think no consumers want this kind of relationship with at least some brands, is ignorance.
But the most successful brands are those that infest your life and which it is almost impossible to detach from. How much money and for how many years have you paid interest to "your" bank, for example? Could you switch to another bank? Have you tried?
I can easily switch banks in a day. Not crypto-heavy either. No loans, insurance, contracts with set date. Are there any other reasons why switching is hard?
Same. The most longstanding brands in my life are literally the ones I forget about.
I absolutely know people who love companies and products. Like LOVE them, they will be put into tears like a child at Christmas, but they're adults. They grew up with them or their family members died and those companies or products bring back those memories. They played their favorite game while eating that food, etc.
You never know. I'm not sure I've ever loved a company exactly, but I've really really liked a product, or sometimes just a type of food. If I like a certain food enough and only a certain company sells it, some of that feeling relates with the company too. Like the company cheers you up, BECAUSE they sell it. You can see how that might be valuable in spreading appeal for the company and helping preserve the thing you enjoy, so it has social/natural selection value.
I think when companies refer to this, they really are referring to real people, it's just aspirational that other people could feel that way if they let themselves. Most people won't.
I think most people can identify a few products they love
If you have a sufficiently broad definition of 'product', then yes.
Millions of people are passionate about brands like the "Dallas Cowboys" and "Star Wars" and will dress up in costumes and go to events with like-minded people.
But for normal products, like USB headsets? Nobody's dressing up as the Jabra 20 Stereo USB-C Headset to go to the big Jabra Convention.
I have brands/products I love, but I love them because I trust them, and typically my trust of something is inversely correlated with seeing mainstream marketing for it.
I agree...but I also think it's a slow path to monetize OpenAI if they don't support some kind of channel to get paid by the sell-side. At least they offer a subscription and didn't jump to this as the only option. I'm optimistic that a premium tier can exist that avoids this, but we'll see. The reality is that they need the money.
> it's a slow path to monetize OpenAI if they don't support some kind of channel to get paid by the sell-side.
Seems like an express route to irrelevance, to me.
If I wanted AI-generated shill reviews, written without laying eyes on the product, hoping to get me to click on an affiliate link? I'd go to Google.
It means the enshittification is reaching full acceleration. The MBAs have taken over. It'll be a constant battle between monetization and research, and a continual skewing of research and development as unnecessary expenses, followed by a full-on torching of the reputation in return for fast cash now and trading on vibes thereafter, while the rest of the world moves on.
Sama wants to speedrun the Apple arc, it looks like.
And investors will be genuinely surprised when the whole thing falls apart in 36 months and the still-very-rich founder spends the next 16 years giving uncomfortably veiny public talks about some uncomfortably sociopathic thing he’s become fascinated with and everyone tries to pretend is totally normal.
It doesnt take an MBA to figure out that selling stuff, or facilitating the sale of stuff, has value. The techies are the ones who built it, the whole "MBA" idea is a coping mechanism to deflect from your own (assuming you're some kind of techie/programmer) participation.
The MBAs have taken over a place that calls itself HackerNews.
Imagine if the company has AI in its name.
This is a brilliant idea. I'm going to mail 99999999 people about my new MBAI degree offering!
For only $6666 you too can become part of the new Mega Banana Annihilation Infantry – the secret task force fighting rogue potassium. Success guaranteed!
Buy now!
Blame Apple, I think they started it decades ago. And yes some people really do love their products.
I think it goes back way farther than that. Look at car ads from the 1950s and 60s, for example. They give off that romantic vibe.
Yeah, miss good ol' mom, gradma talk: you (should) love people; things you like them.
It's true only for a select few products that are actually elite. Like Apple or Dyson.
I have a product I love - it's a barbecue sauce. I buy it from a fella in Texas who ships the bottles in boxes he clearly packs himself (often the bottles have personal thank you notes with my name written on them in sharpie.) Wouldn't call it "elite"
That sounds pretty elite to me. I didn't mean it as elitist.
Tangent to your point, but I think Dyson is generally considered to be overpriced for what it is. Shark and Kenmore are just as good by most metrics and are less expensive, whereas if you want reliable products then you get Miele or Sebo. Dyson's one selling point is that they're typically close to or on the cutting edge of technology, so they're the best if you value noise and size more than price and reliability.
Johnny is frowning right now.
Is this OpenAI's Amazon Dash?
Wonder what protocol is going to win
openai speed running their enshitification
That's the whole point of AI: massive productivity increase so you can compete with the major global tech companies!
Pandora’s box is now open
Enshitification implies there was a good stage to fall from.
this is what NVIDIA invested $100B for!!!
Why would anyone trust the black box to make purchasing decisions and how would that jive with current credit/debit card agreements?
So this is competing with Google's AP2 announced a few weeks ago?
https://cloud.google.com/blog/products/ai-machine-learning/a... (https://news.ycombinator.com/item?id=45262858)
That's what it looks like. This post gives me the impression it is a bit more narrowly scoped and streamlined compared to A2P: https://stripe.com/blog/developing-an-open-standard-for-agen...
This is for the AI engineers who are loaded enough to find it funny that the system orders a lamborghini when they asked for a toy for their kids.
Seems like the best possible "ads-in-chatgpt" implementation. ChatGPT still gives its "Honest Opinion" but you can buy with one click.
(IIUC)
Who, on the user side, is asking for this? Even amazon subscribe is still buying you a product you already chose.
It's a natural step towards agents buying things on your behalf - getting people warmed up to the idea that ChatGPT can be tasked with buying things. It starts with a buy button, "buy this now". In a few weeks it will become "buy dogfood", and it will just do the thing, knowing you get some certain food from Chewy
Who needs that, though? Is buying things that hard and time-consuming?
> In a few weeks it will become "buy dogfood", and it will just do the thing, knowing you get some certain food from Chewy
I have autoship orders set up with Chewy. Stuff arrives on a schedule and I get a small discount. I don't need an overhyped autocomplete fucking things up for me, especially when I can just set up a subscription myself and forget about it.
You're kind of echoing exactly what I'm getting at though. The difference comes down to trust. Maybe today we don't trust that one of these LLM apps is up to the task of buying things, but there was once a time where people didn't trust that if you click a button on a website a product would show up at your door. Over time, we came to expect things like order received emails, status updates, and the best companies will show you where your product is on a little map as it arrives to your house. Now most shopping is done online. Convenience won. There are still barriers to shopping online like too much choice, selecting sizes, and very personal constraints. An agent can know you're in the market for a new phone and can help you pick out a new one, then keep an eye out for deals. Maybe your agent will know your sizes and buy shoes that always fit. It's really not that wild of a concept.
You do not. But the people who eagerly jumped to smart homes, and voice assistants that are always on, to which privacy might as well not exist... Are going to love their own "Jarvis" which they can give shopping orders to while in the kitchen.
Basically, the Sillicon Valley "smart fridge" episode. Or even better, the Modern Family "smart fridge" episode.
Lotsa new stuff coming out of the Department of Who Asked For This Shit these days.
Be quiet peasant and consume when you’re told!
What about ad-blockers? Are they going to lobby browser manufacturers to start enforcing immutable, cryptographically-verified pages?
This will only be profitable is OpenAI makes money from showing the products because the conversion rate will be bad.
Wait until you see the information they’re synthesizing from your conversations and integrations.
Associated site: https://www.agenticcommerce.dev/
Stripe post: https://stripe.com/blog/developing-an-open-standard-for-agen...
all hail. kneel before the altar of capitalism. may our bishop ai shepard us to the light
Holy fuck, wasn't just a few weeks ago someone here who complained about some vibe coding service wiping all the data?
The last thing I'd trust any AI agent, particularly in today's time, would be something tying directly into my bank account!
https://developers.openai.com/commerce
Yet another protocol
Hackernews used to dunk on Airbnb, Coinbase and Uber. Now they're part of our daily lives. Feels like we’re watching the same arc play out with LLMs.
I wonder if we’re seeing the same pattern repeat here with ChatGPT becoming big.
Outrage followed by inevitability.
The problem isn't that OpenAI is big news. The problem is that OpenAI currently appears to be worth more than AirBNB, Coinbase, Uber and Lyft combined, and Sam would have you believe this is still AI's "early stage". How much more liquidity is even left for this thing to soak up?
Yeah, the valuations are crazy, but think about it this way: the app economy and gig economy seemed absurdly speculative at first too. And now they’re worth trillions.
Early stage hype overshoots, sure, but sometimes it’s just pricing in things the old frameworks can’t even model yet.
It's basically the "Oh no! Anyway..." Clarkson meme.
Airbnb and Uber are shit in 2025. They're basically no better than what they were trying to replace. "The arc" you describe is people saying that these products aren't a magical solution that prints money and being proven right as the products are worn down to the level of the thing they were trying to disrupt by market forces.
Uber, in particular, drives me nuts because we replaced a supposedly powerful and evil taxi cartel (which happened to be a bunch of small, regional businesses) with a huge multinational corporation. Out of the frying pan and into the fire.
Totally get the frustration. Uber and Airbnb definitely have their flaws. But the sheer scale of users (200M last I checked) shows that a product can be both polarizing and essential at the same time.
Kind of like how we’re seeing with LLMs: not perfect, often overhyped, but undeniably useful at scale.
Taxis and housing existed and where more afordable before those companies took over the market.
If I make a company buy every hospital in the world you will think is essential, but is not. The hospitals did exist before.
Convenience, consistency, and global. They solved real pain points, even if the cost and ethics are now questionable.
It’s like a software library that’s buggy and bloated but everyone depends on it because rewriting it from scratch is harder than dealing with its flaws.
You start from a false premise, that we have now is better than before.
Taxis were better than Ubers, bed&breakfast were better than AirBnB.
Maybe you are just too young to know it. The only reason we moved to Uber is because it was half the price of taxis, because it was subsidized, not anymore.
Uber and Airbnb solved real pain points: global availability, predictable, nice interfaces, cashless payments, reviews etc. These are the things small, local systems struggled to provide consistently.
Also, my point isn't that the current system is objectively better. It's that scale and convenience created network effects that make it "essential" in practice, even if it’s buggy, slow, or worse in some respects.
Taxis and B&Bs and Hotels all still exist and compete with Uber and AirBnB.
Even at the same price there are valid reasons why many people prefer an Uber over a Taxi, in particular the predictable pricing and globally consistent UI.
Uber changes the pricing depending on the demmand. Taxis can't do that, they are regulated by law in most of the west.
Predictable my ass. You have been lied to.
And btw all over the world you rise up your arm, and the taxi stops, I think that is a pretty consistent user interface that anybody in the world can understand. I have to help my aunt each time she needs an Uber.
Hundreds of people in my circle who use airbnb and uber daily/monthly disagree with you. Could this be unnecessary pessimism?
That you are right won't stop people from being bootlickers.
The parasititic relationships that people from whit mega corporations in the west is so upseting.
This reeks of desperation to generate revenue where they can, with a race against time to show profitability.
The hire of Simo and acquisition of Statsig is very present here.
So who do you call when it drains your account and orders 50 tons of wood, or falls for prompt injection and pays a scammer? Will OpenAI reimbuirse you? No.
Considering how vulnerable AI is to this sort of thing, there's no way in hell I'm touching it for at least 5-10 years.
Sure, they tell you that it's safe at the bottom and give reassurances, but I'm also not going to trust them on that, just as I don't trust pretty much anything ChatGPT tells me without futher verification.
Should have always been clear the singularity will have a shopping cart
But Amazon has a patent on one-click purchases, so with AGI you must ask twice.
I think this is probably a joke, but it’s an incorrect one I believe. Amazon did have such a patent, but it expired in 2017.
Indeed, I believe that the singularity is actually the point in time when the number of shopping carts will tend to infinity.
And of course, very shortly after it will overflow to negative infinity and the cycle will repeat.
And a large billboard blocking the view
AGI will be the ultimate consumer experience.