That makes sense. Probably bad AI summaries. I copied that quote from Wang's page. Looking now at the podcast page [1]:
What are people actually doing with Muse?
Zuckerberg's Muse sources obscure teas weekly from farms in Taiwan and Japan. Wang's favorite use is MMA coaching: cameras in his gym let Muse watch his sparring"
I think this would be a good thing to have, but it has to be comprehensive.
Selective ommission and reduction of context would allow any narrative you chose to write.
In other words for this to work you really need to cover enought scope that it would be surprising to find a public statement that was not covered here.
That's quite a lot of work, but anything else would be susceptible to cherry picking.
YouTube's auto-generated captions are a great product, but one problem is that they can't tell speakers apart in podcasts with multiple people. I'm building something that also works with YouTube transcripts, and it has the same problem.
Speaking nothing of the implementation, this is a very valuable concept. I’m curious if statistically there are trends to identify in beliefs or how beliefs spread over societies. You could likely use that data to determine thought leadership and information diets and measure things like distribution, how far the Bay Area is ahead, and other meaningful data for investments, etc.
Ages ago there was a site called whizbang.com that did something like that on web sites. It tracked various corporate "about" pages and noted how stuff on them changed over time. It may have done other stuff too. It was almost unusably slow and shut down after a while. I assume someone else has the domain now and is doing something unrelated with it.
I have a friend that’s working on a claim provenance database, starting with a focus on Wikipedia and news sources but ideally it could be applied to social media fire hoses as well.
Hmm I feel like the social media aspect would be way more useful for democracies in general.
I remember reading about Switzerland’s government service that tracks each politician, their political actions - which positions they take, which laws they promote, etc. and comparing that with which companies and investments they, their famalies and their close connections have.
All quantifiable data. I always thought something like this was _essential_ for any democracy in general - follow the money and “qui bono” and all that, thought it would be cool to build one. But of course prohibitively expensive for individuals small teams. But with all the ai - this could definitely be a thing.
If it all boils down to “verifiable corruption rating” as a single sound bite, it could seriously affect public opinion I feel. Hope someone gets on that idea someday.
There used to be a show in Argentinean TV which weaponized this idea. The strategy was to take something a politician said that week and contrast it with a videoclip (months, years, even decades old) of that politician saying (apparently) the opposite.
It's not obvious at first, but this idea is actually pretty terrible: changing your mind in presence of evidence is something we want politicians to do, but if you then use that to paint them as hypocrites then they'll never ever change any stated position. And "the opposite" casts a wide net: if someone is recorded as saying "I think we should stop building datacenters for AI" and you then contrast them with an old clip saying they use Google Translate, what does that tell us about the character of this person? (IMO nothing, but that's not what the clip wants to tells you)
This is...kind of my fear though. We need some pragmatism in politics but the American public fucking hates anything other than "my team best team forever". I say this as someone who is staunchly more left than probably 95% of everyone here... Building an algorithm that ensures absolutely any deviation from the trend is swiftly punished brutally is not necessarily a feature of a functional society.
You may think you want this because of Sinema and Fetterman but...we already know they're shitheads because...they blatantly do shithead stuff. Replacing the time spent gleaning all the nuance behind a vote with a party whip algorithm is not going to somehow make voters better at voting.
No they don’t - they can ‘influence’ and hope to get their agenda pushed through but no guarantee unless the politician is desperate and you’re putting forth a lot of money (aka musk when trump got re elected).
Show me how often there were lasting decisions made against one of the 'fat cats' and their companies other than to bring them to heel.
Money and politics should never mix. One man, one vote. Not one dollar, one vote. One E.M. just tossed a few hundred million on the table in order to influence the outcome of an election and nobody did anything about it. That should tell you enough.
It's hard to see how ending John Stumpf's career, for example, was about bringing anyone to heel. It seems like a clear and unambiguous example of Congress and federal regulators realizing that Wells Fargo had done something bad and punishing both the company and CEO for it. If you want examples of someone going to prison, there's Sam Bankman-Fried and Elizabeth Holmes, although I could imagine arguments that they don't count as "fat cats".
didnt he just shell out another extra 276 mil (accountable donations) and several other in non-accountable donations (allegedly)? he probably still pissed that peter thiel managed to place his man as vp for peanuts while he got the sack after only six months and 100mil spent. fun to watch caligula and his oligarch bidders taking apart the empire bit by bit.
> fun to watch caligula and his oligarch bidders taking apart the empire bit by bit.
It'd be fun if so many people weren't being hurt in the process. A part of me can empathize with the folks who are so disillusioned that they're willing to let the country burn, but we're all still in here...
I've been building a tool that turns YouTube videos and podcasts into searchable transcripts. But somewhere along the way, I became more interested in a different problem.
How do you keep track of what influential people actually say, and how their thinking changes over time?
Take Sam Altman, for example.
In October 2025, talking about compute, he said, "there will come a glut for sure."
In June 2026, he said, "there will be a shortage forever."
Two statements, eight months apart. There may be perfectly reasonable context behind both, but I found it interesting that there wasn't an easy way to see them together, with the original recordings.
It currently follows 20 people in AI, including Sam Altman, Dario Amodei, Alex Karp and Marc Andreessen.
Every day, it looks for new interviews, podcasts and talks where these people are actually speaking, rather than videos of other people talking about them.
For each appearance, you get a short summary, key quotes and links to the exact moments in the original recording.
The feature I'm most excited about is *"What changed".*
It compares what someone says today with their earlier statements on the same topic. You can see both quotes, their dates and the original sources.
I'm not trying to catch people contradicting themselves. People should change their minds when new evidence emerges. I just think it would be useful to have a record of how their thinking evolves.
Building it has been more challenging than I expected.
A surprising number of videos with someone's name in the title don't actually feature them speaking. Reuploads are everywhere. And getting AI to reliably attribute statements to the right person is harder than simply generating a transcript.
I've built filters to handle much of this, and every published "What changed" comparison is still manually reviewed.
It's far from perfect. Speaker identification needs work, some podcasts aren't transcribed automatically, and I'm starting with English only.
On the economics side, a transcript and summary currently cost roughly one US cent in model processing. That makes tracking a fairly large history surprisingly affordable.
Everything is free to read. You can also follow up to three people, receive updates when they speak, and ask questions like "What has Dario Amodei said about open-source AI recently?" with answers linked to the original recordings.
I'm still figuring out which part is genuinely valuable: keeping up with people, searching their past statements, or spotting changes in their views.
I'd love some feedback from HN.
Who else should I track? Would you actually use the "What changed" feature? And if you find something incorrectly attributed or summarised, please report it. I'd rather find the failures early.
There was an interview with Altman where he was talking about various regulatory policies that had been enacted, then at the end he added "but in practice none of it turned out to matter. What we didn't anticipate was that the companies became more powerful than the governments". I wish I could find it today but it seems gone
Congrats on building this Gaurav. I think you can take this a step further and create how alignment is formed in powerful people's views (in public statements, at least) on important topics over time, and how potential political and world events drive those alignments.
Alexandr Wang "says his favorite Muse use case is having it coach him in MMA by watching gym cameras and giving feedback between rounds"
Okay, you've convinced me I don't want to track what these guys are saying.
On a serious note, it looks good. Although to really track what they say, you'd probably want text too, e.g tweets and blog posts.
Wasn't that Zuckerberg? And Alexandr Wang saying he uses Muse to buy obscures teas from Japan and Taiwan?
That makes sense. Probably bad AI summaries. I copied that quote from Wang's page. Looking now at the podcast page [1]:
What are people actually doing with Muse?
Zuckerberg's Muse sources obscure teas weekly from farms in Taiwan and Japan. Wang's favorite use is MMA coaching: cameras in his gym let Muse watch his sparring"
[1] https://blumify.io/t/youtube/XfdIhd1g6EI/muse-wonder-zuckerb...
I think this would be a good thing to have, but it has to be comprehensive.
Selective ommission and reduction of context would allow any narrative you chose to write.
In other words for this to work you really need to cover enought scope that it would be surprising to find a public statement that was not covered here.
That's quite a lot of work, but anything else would be susceptible to cherry picking.
Thanks for your feedback Lerc, yes it is a lot of work, but I wanted to ask the community if it is valuable enough to start with.
YouTube's auto-generated captions are a great product, but one problem is that they can't tell speakers apart in podcasts with multiple people. I'm building something that also works with YouTube transcripts, and it has the same problem.
It’s called “diarization“ and the leading models are not nearly as accurate as the leading speech to text - whisperx etc.
You could use some smaller models, like parakeet, to figure out the users.
Thanks for the feedback.
Karp's takes and changes in tune are... depressing.
Interesting concept but useless without signing up for who knows what just to start trying to see what things all about.
Speaking nothing of the implementation, this is a very valuable concept. I’m curious if statistically there are trends to identify in beliefs or how beliefs spread over societies. You could likely use that data to determine thought leadership and information diets and measure things like distribution, how far the Bay Area is ahead, and other meaningful data for investments, etc.
Ages ago there was a site called whizbang.com that did something like that on web sites. It tracked various corporate "about" pages and noted how stuff on them changed over time. It may have done other stuff too. It was almost unusably slow and shut down after a while. I assume someone else has the domain now and is doing something unrelated with it.
ahead?
Negative numbers are allowed ;)
You noticed that too, eh?
Lol i m not from silicon valley.
I have a friend that’s working on a claim provenance database, starting with a focus on Wikipedia and news sources but ideally it could be applied to social media fire hoses as well.
Hmm I feel like the social media aspect would be way more useful for democracies in general.
I remember reading about Switzerland’s government service that tracks each politician, their political actions - which positions they take, which laws they promote, etc. and comparing that with which companies and investments they, their famalies and their close connections have.
All quantifiable data. I always thought something like this was _essential_ for any democracy in general - follow the money and “qui bono” and all that, thought it would be cool to build one. But of course prohibitively expensive for individuals small teams. But with all the ai - this could definitely be a thing.
If it all boils down to “verifiable corruption rating” as a single sound bite, it could seriously affect public opinion I feel. Hope someone gets on that idea someday.
thank you for your feedback, I will definitely look into it.
Quoting a Kai Lentit video as Dario Amodei changing his mind on regulation is peak slop.
Cool idea, I guess, but its just more slop to track the slop.
Indeed. This is not Dario Amodei, despite the caption...
https://www.youtube.com/watch?v=DlTNN0gvkLM
Very funny, btw. "Can we please stop anthropomorphising the LLMs? They hate that."
done..
I wanted to get a place where I can see get info. correctly and AI leaders changing their mind is a slop in itself.
"AI can make mistakes".
super nice idea , should do that on politics
There used to be a show in Argentinean TV which weaponized this idea. The strategy was to take something a politician said that week and contrast it with a videoclip (months, years, even decades old) of that politician saying (apparently) the opposite.
It's not obvious at first, but this idea is actually pretty terrible: changing your mind in presence of evidence is something we want politicians to do, but if you then use that to paint them as hypocrites then they'll never ever change any stated position. And "the opposite" casts a wide net: if someone is recorded as saying "I think we should stop building datacenters for AI" and you then contrast them with an old clip saying they use Google Translate, what does that tell us about the character of this person? (IMO nothing, but that's not what the clip wants to tells you)
This is...kind of my fear though. We need some pragmatism in politics but the American public fucking hates anything other than "my team best team forever". I say this as someone who is staunchly more left than probably 95% of everyone here... Building an algorithm that ensures absolutely any deviation from the trend is swiftly punished brutally is not necessarily a feature of a functional society.
You may think you want this because of Sinema and Fetterman but...we already know they're shitheads because...they blatantly do shithead stuff. Replacing the time spent gleaning all the nuance behind a vote with a party whip algorithm is not going to somehow make voters better at voting.
thanks for the suggestion mate.
AI leaders aren’t important, why not track actual politicians?
Technically, they own the politicians, or at least a substantial fraction of them.
No they don’t - they can ‘influence’ and hope to get their agenda pushed through but no guarantee unless the politician is desperate and you’re putting forth a lot of money (aka musk when trump got re elected).
Show me how often there were lasting decisions made against one of the 'fat cats' and their companies other than to bring them to heel.
Money and politics should never mix. One man, one vote. Not one dollar, one vote. One E.M. just tossed a few hundred million on the table in order to influence the outcome of an election and nobody did anything about it. That should tell you enough.
It's hard to see how ending John Stumpf's career, for example, was about bringing anyone to heel. It seems like a clear and unambiguous example of Congress and federal regulators realizing that Wells Fargo had done something bad and punishing both the company and CEO for it. If you want examples of someone going to prison, there's Sam Bankman-Fried and Elizabeth Holmes, although I could imagine arguments that they don't count as "fat cats".
didnt he just shell out another extra 276 mil (accountable donations) and several other in non-accountable donations (allegedly)? he probably still pissed that peter thiel managed to place his man as vp for peanuts while he got the sack after only six months and 100mil spent. fun to watch caligula and his oligarch bidders taking apart the empire bit by bit.
> fun to watch caligula and his oligarch bidders taking apart the empire bit by bit.
It'd be fun if so many people weren't being hurt in the process. A part of me can empathize with the folks who are so disillusioned that they're willing to let the country burn, but we're all still in here...
AI leaders are the employers, politicians are just the employees. I don't track my starbucks barista either.
They are politicians, just not elected ones
We should point blumify also at all HN commenters, and I'd expect this to be an opinion that'd change in the next few years :)
That is the plan pls upvote in the website or suggest names there, I will try to add them
Hi HN, I'm Gaurav.
I've been building a tool that turns YouTube videos and podcasts into searchable transcripts. But somewhere along the way, I became more interested in a different problem.
How do you keep track of what influential people actually say, and how their thinking changes over time?
Take Sam Altman, for example.
In October 2025, talking about compute, he said, "there will come a glut for sure."
In June 2026, he said, "there will be a shortage forever."
Two statements, eight months apart. There may be perfectly reasonable context behind both, but I found it interesting that there wasn't an easy way to see them together, with the original recordings.
It currently follows 20 people in AI, including Sam Altman, Dario Amodei, Alex Karp and Marc Andreessen.
Every day, it looks for new interviews, podcasts and talks where these people are actually speaking, rather than videos of other people talking about them.
For each appearance, you get a short summary, key quotes and links to the exact moments in the original recording.
The feature I'm most excited about is *"What changed".*
It compares what someone says today with their earlier statements on the same topic. You can see both quotes, their dates and the original sources.
I'm not trying to catch people contradicting themselves. People should change their minds when new evidence emerges. I just think it would be useful to have a record of how their thinking evolves.
Building it has been more challenging than I expected.
A surprising number of videos with someone's name in the title don't actually feature them speaking. Reuploads are everywhere. And getting AI to reliably attribute statements to the right person is harder than simply generating a transcript.
I've built filters to handle much of this, and every published "What changed" comparison is still manually reviewed.
It's far from perfect. Speaker identification needs work, some podcasts aren't transcribed automatically, and I'm starting with English only.
On the economics side, a transcript and summary currently cost roughly one US cent in model processing. That makes tracking a fairly large history surprisingly affordable.
Everything is free to read. You can also follow up to three people, receive updates when they speak, and ask questions like "What has Dario Amodei said about open-source AI recently?" with answers linked to the original recordings.
I'm still figuring out which part is genuinely valuable: keeping up with people, searching their past statements, or spotting changes in their views.
I'd love some feedback from HN.
Who else should I track? Would you actually use the "What changed" feature? And if you find something incorrectly attributed or summarised, please report it. I'd rather find the failures early.
There was an interview with Altman where he was talking about various regulatory policies that had been enacted, then at the end he added "but in practice none of it turned out to matter. What we didn't anticipate was that the companies became more powerful than the governments". I wish I could find it today but it seems gone
let me find that and post it
Congrats on building this Gaurav. I think you can take this a step further and create how alignment is formed in powerful people's views (in public statements, at least) on important topics over time, and how potential political and world events drive those alignments.
thanks mate.
Great problem to work on!
thanks