It's been about 15 years since I've worked with RoR, but my favorite aspect of ruby was and will always be the library names. Shout out to factory_girl which I found out this morning was unforunately renamed to factory_bot
At least this is the end of the js framework of the week. We developers deliberated ourselves from software development. Hell. We made our own jobs redundant. How stupid and genius at the same time can a profession be
I imagine the perfect programing language would have 3 levels.
1.LLM "code" , this should work for most basic use cases.
Should be so basic any random person can create a CRUD app.
2. Scripting, something like Python. This should handle 95% of use cases.
3. Systems programing. Zig, Rust, etc. For when you need extremely specific performance requirements to be met.
My dream language would integrate all three of these in the same stack, ideally the same project would be a mix of all three ( most of the time a mix of the first two).
While “silly” this is likely the next paradigm/abstraction for intent based pages.
You can imagine given 1,000,000 page views just how many experiments could be run. Basically our A/B tests start to resemble natural evolution and survival of the fittest more than decision trees.
However, something feels like it’s missing. I wonder what’s still yet to be built before we arrive at that future.
Yeah, totally agree that something related to this will likely be the next paradigm. I've been putting together experiments in different directions trying to find that thing that's missing but haven't really found a killer use case yet to pull it all together.
That's a really cool idea that once you can get something somewhat reliably consistent generated, you can kind of let your A/B tests start to run themselves with just rough guidelines on what you're trying to optimize for...
> Every VP thinks their idea for a feature will revolutionize the company.
Now imagine that everyone of them is given a tool that could get them an POC quickly. I think a lot VPs are about to figure out that their ideas are shit.
I love that this is more art piece than serious software... more like offering someone an expedition than a product.
(Though I'm not sure I'll get on the expedition, I am a little worried about sandboxing and setup and getting distracted...)
If I was to start the expedition, I'd probably try to overshoot by describing a site that I could not myself fully imagine, or using attributes that lacked a single meaning. Like, "the artist's interactive portfolio, as though the artist is looking over your shoulder, the artist keeping a carefully neutral expression while seething inside." Then I'd probably continue, imagining just the outline of some site that satisfies some unarticulated desire, putzing around as I see a concrete articulation of that idea, as much reforming the idea in my head in response to those results to an equal degree that I am articulating the idea in more detail.
Ahh I absolutely love this idea of trying to infuse more emotional and fuzzy attributes to see what the LLM comes up with!
When I broke out the layout and style components I was thinking of being able to change the whole site aesthetic from something like "standard b2b" to "geocities fan page", but I'm excited to try getting fuzzier with the descriptions!
It's literally a nod to the fact that the framework may give you something close to what you want but off in some way. It's pretty much the perfect name for what it is.
My very first thought was similar, followed by recalling a ruby whitespace issue that treated the non-space-whitespace as an undefined function. That was harder to debug than it should have been.
Instead, after reading the page, it is LLM generated pages where "you get what you ask for," hallucinations and all. Fantastic name.
I have always hated Gherkin because the extra layer of language expression / abstraction is superficial, and PMs can't write any ol' thing that makes sense, it has to be supported by the parser.
But LLMs can make sense of any ol' thing, so, and it shocks me to admit such, maybe Gherkin is back on the menu.
I also created an experiment for this, giving the AI an ability to write/read from a database so you could build full CRUD apps.
It works somewhat but even with the smaller/faster models it's very slow and even with the big models it is pretty unreliable. Long term I can definitely imagine this will get more viable and maybe become a complement to the 'chat' interface with most SaaS apps essentially being replaced with a AI in front of system or systems of record.
It's been about 15 years since I've worked with RoR, but my favorite aspect of ruby was and will always be the library names. Shout out to factory_girl which I found out this morning was unforunately renamed to factory_bot
At least this is the end of the js framework of the week. We developers deliberated ourselves from software development. Hell. We made our own jobs redundant. How stupid and genius at the same time can a profession be
I imagine the perfect programing language would have 3 levels.
1.LLM "code" , this should work for most basic use cases. Should be so basic any random person can create a CRUD app.
2. Scripting, something like Python. This should handle 95% of use cases.
3. Systems programing. Zig, Rust, etc. For when you need extremely specific performance requirements to be met.
My dream language would integrate all three of these in the same stack, ideally the same project would be a mix of all three ( most of the time a mix of the first two).
While “silly” this is likely the next paradigm/abstraction for intent based pages.
You can imagine given 1,000,000 page views just how many experiments could be run. Basically our A/B tests start to resemble natural evolution and survival of the fittest more than decision trees.
However, something feels like it’s missing. I wonder what’s still yet to be built before we arrive at that future.
Yeah, totally agree that something related to this will likely be the next paradigm. I've been putting together experiments in different directions trying to find that thing that's missing but haven't really found a killer use case yet to pull it all together.
That's a really cool idea that once you can get something somewhat reliably consistent generated, you can kind of let your A/B tests start to run themselves with just rough guidelines on what you're trying to optimize for...
I think that’d make for interesting experiments and fringe sites, I don’t really see like your average e-commerce site ever doing anything like that.
You’d want the A and B to be intentional, not automatically generated. Every VP thinks their idea for a feature will revolutionize the company.
> Every VP thinks their idea for a feature will revolutionize the company.
Now imagine that everyone of them is given a tool that could get them an POC quickly. I think a lot VPs are about to figure out that their ideas are shit.
This pre-supposes that said VPs have the self-awareness to realise their ideas are shit.
Well, you missed the bit where ad and marketing networks get involved and corrupt it into some god-awful granular targeting system.
yeah, sorry, I don't want optimal dark patterns.
We got enough of that atm.
This is it folks, we have finally reached peak of webdev
Wow, cool to see this make it on to HN!
Author here, happy to answer any questions about this or chat about the ideas behind it :)
I love that this is more art piece than serious software... more like offering someone an expedition than a product.
(Though I'm not sure I'll get on the expedition, I am a little worried about sandboxing and setup and getting distracted...)
If I was to start the expedition, I'd probably try to overshoot by describing a site that I could not myself fully imagine, or using attributes that lacked a single meaning. Like, "the artist's interactive portfolio, as though the artist is looking over your shoulder, the artist keeping a carefully neutral expression while seething inside." Then I'd probably continue, imagining just the outline of some site that satisfies some unarticulated desire, putzing around as I see a concrete articulation of that idea, as much reforming the idea in my head in response to those results to an equal degree that I am articulating the idea in more detail.
Ahh I absolutely love this idea of trying to infuse more emotional and fuzzy attributes to see what the LLM comes up with!
When I broke out the layout and style components I was thinking of being able to change the whole site aesthetic from something like "standard b2b" to "geocities fan page", but I'm excited to try getting fuzzier with the descriptions!
This reminds me of something _why the lucky stiff would create.
I don't know if I would use a framework named after something that's by nature unreliable, or even devious.
It's literally a nod to the fact that the framework may give you something close to what you want but off in some way. It's pretty much the perfect name for what it is.
It’s clearly on purpose. The tagline on the page is “Be careful what you wish for...”
It's only an issue if you get a copy by cURL-ing it, dohohoho.
My very first thought was similar, followed by recalling a ruby whitespace issue that treated the non-space-whitespace as an undefined function. That was harder to debug than it should have been.
Instead, after reading the page, it is LLM generated pages where "you get what you ask for," hallucinations and all. Fantastic name.
> Natural language as source code: Your intention becomes the program
Reminds me of Cucumber testing framework
Interestingly, Cucumber (with its Gherkin syntax) works quite well for detailed and structured prompting to LLMs.
I have always hated Gherkin because the extra layer of language expression / abstraction is superficial, and PMs can't write any ol' thing that makes sense, it has to be supported by the parser.
But LLMs can make sense of any ol' thing, so, and it shocks me to admit such, maybe Gherkin is back on the menu.
I also created an experiment for this, giving the AI an ability to write/read from a database so you could build full CRUD apps.
It works somewhat but even with the smaller/faster models it's very slow and even with the big models it is pretty unreliable. Long term I can definitely imagine this will get more viable and maybe become a complement to the 'chat' interface with most SaaS apps essentially being replaced with a AI in front of system or systems of record.