One small one I do not agree with is "Are you burning VC cash on unsustainable unit economics?". I think it's safe to conclude by now that unsustainable businesses can be kept alive for years as long as the investors want it.
It is becoming unbearable. YouTube now has "AI" slop ads for Freenow (Lyft brand in the EU) with fake cars that move without the wheels turning and "AI" "actors" that look like plastic.
This of course means that Freenow is now on the personal blacklist. People should not engage with companies who advertise with "AI" slop.
Doesn't matter. We must keep building more and more technology no matter the cost. Have an idea for a business? Build it. Does your business make the lives of people worse? Doesn't matter, keep pushing. Could some new technology ruin the lives and relationships that people have? Doesn't matter, just build it. We always need more, need to do more. Every experiment is valid, every impulse must be followed. More complexity, more control, more distraction, more outrage, more engagement. Just keep building forever no matter the cost.
Yuval Noah Harari, Sapiens fame [0], has a great quote (paraphrasing):
Interviewer: How will humans deal with the avalanche of fake information that AI could bring?
YNH: The way humans have always dealt with fake information: by building institutions we trust to provide accurate information. This is not a new phenomenon btw.
In democracies, this is often either the government (e.g. the Bureau of Labor Statistics) or newspapers (e.g. the New York Times) or even individuals (e.g. Walter Cronkite).
In other forms of government, it becomes trust networks built on familial ties e.g. "Uncle/Aunt is the source for any good info on what's happening in the company" etc
> YNH: The way humans have always dealt with fake information: by building institutions we trust to provide accurate information. This is not a new phenomenon btw.
Funny that he doesn’t say that the institutions have to provide accurate information, but just that we have to trust them to provide accurate information.
AI-esque blog post about how infinite AI content is awful, from "a co-founder at Paid, which is the first and only monetization and billing system for AI Agents".
I needed to get some builder quotes for my home. It did not enter my mind to go online to search for any.
I just reached out to my family for any trustworthy builders they've had, and struck up conversations with some of my fancier neighbors for any recommendations.
(I came to the conclusion that all builders are cowboys, and I might as well just try doing some of this myself via youtube videos)
Using the internet to buy products is not a problem for me, I know roughly the quality of what I expect to get and can return anything not up to standard. Using the internet to buy services though? Not a chance. How can you refund a service
What I get from the article is that, proving that a company will stick around for a while after you’ve subscribed is hard now, because anybody can AI generate the general vibe of the marketing department of a big established player. This seems like it’ll be devastating for companies whose business model requires signing new users up for ongoing subscriptions.
Maybe it could lead to a resurgence of the business model where you buy a program and don’t have to get married to the company that supports it, though?
I’d love it if the business model of “buy our buggy product now, we’ll maybe patch it later” died.
What you’re describing is basically the Drift Principle. Once a system optimizes faster than it can preserve context, fidelity is the first thing to go. AI made the cost of content and the cost of looking credible basically zero, so everything converges into the same synthetic pattern.
That’s why we’re seeing so much semantic drift too. The forms of credibility survive, but the intent behind them doesn’t. The system works, but the texture that signals real humans evaporates. That’s the trust collapse. Over optimized sameness drowning out the few cues we used to rely on.
I'm already seeing this. I very much fall into the category of 'delete all email offers' as I'm a small youtuber, big enough to be targeted by AI sponsor deals, so I'm just buried with it.
The last five times I've looked at something in case it was a legitimate user email it was AI promotion of someone just like in the article.
Their only way to escalate, apart from pure volume, is to take pains to intentionally emulate the signals of someone who's a legitimate user needing help or having a complaint. Logically, if you want to pursue the adversarial nature of this farther, the AIs will have to be trained to study up and mimic the dialogue trees of legitimate users needing support, only to introduce their promotion after I've done several exchanges of seemingly legitimate support work, in the guise of a friend and happy customer. All pretend, to get to the pitch. AI's already capable of this if directed adeptly enough. You could write a script for it by asking AI for a script to do exactly this social exploit.
By then I'll be locked in a room that's also a Faraday cage, poking products through a slot in the door—and mocking my captors with the em-dashes I used back when I was one of the people THEY learned em-dashes from.
One thing about it, it's a very modern sort of dystopia!
Trust collapse is real, I don't trust anything anymore. Take this article for instance, I don't trust it because of the random bolding. Does that mean it's AI generated? I don't know but I've seen lots of AI generated content and it has random bolding, so when I see it, I immediately don't trust it. And I don't have the time to verify anything, so whether or not this article was written by the author or AI, it's gone on the "not credible" heap for me, just because of the bolding. It's not a strong signal but it's a signal, and due to the volume of slop, I must filter on whatever signals I have to maintain any chance of finding genuine human work product. Maybe I miss something genuine and important by filtering this way, but it's the best I can do.
> Will you still be here in 12 months when I’ve integrated your tool into my workflow?
This is the biggie; especially with B2B. It's really 3 months, these days. Many companies have the lifespan of a mayfly.
AI isn't the new reason for this. It's been getting worse and worse, in the last few years, as people have been selling companies; not products, but AI will accelerate the race to the bottom. One of the things that AI has afforded, is that the lowest-tier, bottom-feeding scammer, can now look every bit as polished and professional as a Fortune 50 company (often, even more).
So that means that not only is the SNR dropping, the "noise" is now a lot riskier and uglier.
You can't build trust in your OS (operating system) when your OS spies on the entire customer base, and you spin it off as telemetry. Or you remotely target the OS to implement a radical change, and force it to be installed as an 'update'.
I stopped accepting telephone calls before 2010. They still ring the phone.
What if this is the plan all along? People losing trust in media, so the rich and powerful can continue doing shit without getting exposed any more, because now they always can say it's just AI, and didn't really do this or that?
I wish we were talking about what's next versus what's increasingly here.
How can infinite AI content be strictly awful if it forces us to fix issues with our trust and reward systems? Short term, sure. But infinite (also) implies long term.
I wish I had a really smart game theorist friend who could help me project forward into time if for nothing other than just fun.
Don't get me wrong, I'm not trying to reduce the value of "ouch, it hurts right now" stories and responses.
But damned if we don't have an interesting and engaging problem on our hands right now. There's got to be some people out there who love digging in to complicated problems.
What's next after trust collapses? All of us just give up? What if that collapse is sooner than we thought; can we think about the fun problem now?
We see business go through this cycle a lot. Some new “better cheaper” thing comes along. Everyone implements it to keep up with the Jones’s. Suddenly there’s no differentiation because everyone has it and everyone thinks it sucks. Suddenly going back to some reworked version of the old thing is the new black.
One such example was call centers. In the 2000s implementing a call center in India was all the rage on cost cutting. The customer experience was terrible and suddenly having a US-based call center (the thing companies just abandoned) was now a feature.
I think we’ll see similar things with AI. Everyone will get flooded with AI slop. Folks will get annoyed and suddenly interacting with a real human or a real human writing original content will be a “feature” that folks flock to.
We need PageRank like algorithm for "Trust / Human Content" to be applied directly to the source of such content. E.g. following all three channels are AI made. But all these content can be liked to an advanced AI version of audio based videos of Wikiarticles. If a video is providing just a summary based on established historical facts, even though it is AI based, how is it different than refering a thesaurus or dictionary? Aren't such videos making "knowledge" accessible.
FINAL Financial hours of U.S.A. just before the 1929 crash
It didn't help documentation at all. I had to work with auth0 for example and their documentation is such a bloat, that I am already prototyping with better-auth.
No structure, outdated stuff marked as "preview" from 2023/2024, wikipedia like in depth articles about everything but not for simple questions like: how to implement a backend for frontend.
You find fragments and pieces of information here and there - but no guidance at all. Settings hidden behind tabs etc.
A nightmare.
No sane developer would have done such a mess, because of time constraints and bloat. You see and experience first hand, that the few gems are from the trenches, with spelling mistakes etc.
So I went on X after a long break from social media, and my feed is full of tips like this one:
Growing on X is so simple I’m shocked it works.
100x comments a day
10x posts a day
15x DM’s a day
1x thread a day
1x email a day
This is how you grow your presence on X.
Even if having a presence matters, how can you actually say something meaningful if you post 10 times a day - there's no way (unless you just repeat yourself). Hopefully my algorithm's just gone weird but sadly the people I used to follow stopped posting.
I don't think this shows that you can't trust things. I think it means trust should be earned.
We might be transitioning to a world where trust has value and is earned and stored in your reputation. Clickbait is a symptom of people valuing attention over trust. Clickbait spends a percentage of their reputation by trading it for attention.
In a world of many providers, most people have not heard of any particular individual provider. This means they have no reputation to lose, so their choice to act in a reputation losing manner is easy.
Beyond a certain scale when everyone can play that game we end up with the problem that this article describes. The content is easy but vacuous. There are far more people vying for the same number of eyballs now.
The solution is, I believe, earned trust. Curators select items from sources they trust. The ones that do a good job become trusted curators. In a sense HackerNews is a trusted curator. Reddit is one that is losing, or has lost, trust.
AI could probably take on some of the role of that curation. In the future perhaps more so. An AI can scan the sources of an article to see if the sources make the claims that the article says it makes. I doubt it can do so with sufficient accuracy to be useful right now, but I don't think that is too far off.
Perhaps the various fediverse reddit clones had the wrong idea. Maybe they should in a distributed fashion where each point is a subreddit analogue operated each with their own ways of curation, then an upper level curation can make a site of the groups they trust.
This makes a multi level trust mechanism. At each level there are no rules governing behaviour. If you violate the values of a higher layer, they lose trust in you. AI could run its own curation nodes. It might be good at it or it might be terrible, it doesn't really matter. If it is consistently good, it earns trust.
I don't mind there being lots of stuff, if I can still find the good stuff.
This isn't limited to sales. The trust collapse is also coming for the public debate, interpersonal relationships and probably more stuff than I can imagine right now.
I predict a renaissance of meeting people in person.
41 comments
[ 1.9 ms ] story [ 47.5 ms ] threadOne small one I do not agree with is "Are you burning VC cash on unsustainable unit economics?". I think it's safe to conclude by now that unsustainable businesses can be kept alive for years as long as the investors want it.
This of course means that Freenow is now on the personal blacklist. People should not engage with companies who advertise with "AI" slop.
Interviewer: How will humans deal with the avalanche of fake information that AI could bring?
YNH: The way humans have always dealt with fake information: by building institutions we trust to provide accurate information. This is not a new phenomenon btw.
In democracies, this is often either the government (e.g. the Bureau of Labor Statistics) or newspapers (e.g. the New York Times) or even individuals (e.g. Walter Cronkite).
In other forms of government, it becomes trust networks built on familial ties e.g. "Uncle/Aunt is the source for any good info on what's happening in the company" etc
0 - https://amzn.to/4nFuG7C
Funny that he doesn’t say that the institutions have to provide accurate information, but just that we have to trust them to provide accurate information.
I just reached out to my family for any trustworthy builders they've had, and struck up conversations with some of my fancier neighbors for any recommendations.
(I came to the conclusion that all builders are cowboys, and I might as well just try doing some of this myself via youtube videos)
Using the internet to buy products is not a problem for me, I know roughly the quality of what I expect to get and can return anything not up to standard. Using the internet to buy services though? Not a chance. How can you refund a service
Maybe it could lead to a resurgence of the business model where you buy a program and don’t have to get married to the company that supports it, though?
I’d love it if the business model of “buy our buggy product now, we’ll maybe patch it later” died.
That’s why we’re seeing so much semantic drift too. The forms of credibility survive, but the intent behind them doesn’t. The system works, but the texture that signals real humans evaporates. That’s the trust collapse. Over optimized sameness drowning out the few cues we used to rely on.
I follow even AI slop via reddit RSS.
I control however what comes in.
The last five times I've looked at something in case it was a legitimate user email it was AI promotion of someone just like in the article.
Their only way to escalate, apart from pure volume, is to take pains to intentionally emulate the signals of someone who's a legitimate user needing help or having a complaint. Logically, if you want to pursue the adversarial nature of this farther, the AIs will have to be trained to study up and mimic the dialogue trees of legitimate users needing support, only to introduce their promotion after I've done several exchanges of seemingly legitimate support work, in the guise of a friend and happy customer. All pretend, to get to the pitch. AI's already capable of this if directed adeptly enough. You could write a script for it by asking AI for a script to do exactly this social exploit.
By then I'll be locked in a room that's also a Faraday cage, poking products through a slot in the door—and mocking my captors with the em-dashes I used back when I was one of the people THEY learned em-dashes from.
One thing about it, it's a very modern sort of dystopia!
This is the biggie; especially with B2B. It's really 3 months, these days. Many companies have the lifespan of a mayfly.
AI isn't the new reason for this. It's been getting worse and worse, in the last few years, as people have been selling companies; not products, but AI will accelerate the race to the bottom. One of the things that AI has afforded, is that the lowest-tier, bottom-feeding scammer, can now look every bit as polished and professional as a Fortune 50 company (often, even more).
So that means that not only is the SNR dropping, the "noise" is now a lot riskier and uglier.
I stopped accepting telephone calls before 2010. They still ring the phone.
How can infinite AI content be strictly awful if it forces us to fix issues with our trust and reward systems? Short term, sure. But infinite (also) implies long term.
I wish I had a really smart game theorist friend who could help me project forward into time if for nothing other than just fun.
Don't get me wrong, I'm not trying to reduce the value of "ouch, it hurts right now" stories and responses.
But damned if we don't have an interesting and engaging problem on our hands right now. There's got to be some people out there who love digging in to complicated problems.
What's next after trust collapses? All of us just give up? What if that collapse is sooner than we thought; can we think about the fun problem now?
One such example was call centers. In the 2000s implementing a call center in India was all the rage on cost cutting. The customer experience was terrible and suddenly having a US-based call center (the thing companies just abandoned) was now a feature.
I think we’ll see similar things with AI. Everyone will get flooded with AI slop. Folks will get annoyed and suddenly interacting with a real human or a real human writing original content will be a “feature” that folks flock to.
FINAL Financial hours of U.S.A. just before the 1929 crash
https://www.youtube.com/watch?v=dxiSOlvKUlA&t=1008s
The Volcker Shock: When the Fed Broke the Economy to Save the Dollar (1980)
https://www.youtube.com/watch?v=cTvgL2XtHsw
How Inflation Makes the Rich Richer
https://www.youtube.com/watch?v=WDnlYQsbQ_c
No structure, outdated stuff marked as "preview" from 2023/2024, wikipedia like in depth articles about everything but not for simple questions like: how to implement a backend for frontend.
You find fragments and pieces of information here and there - but no guidance at all. Settings hidden behind tabs etc.
A nightmare.
No sane developer would have done such a mess, because of time constraints and bloat. You see and experience first hand, that the few gems are from the trenches, with spelling mistakes etc.
Bloat for SEO, the mess for devs.
Growing on X is so simple I’m shocked it works.
100x comments a day
10x posts a day
15x DM’s a day
1x thread a day
1x email a day
This is how you grow your presence on X.
Even if having a presence matters, how can you actually say something meaningful if you post 10 times a day - there's no way (unless you just repeat yourself). Hopefully my algorithm's just gone weird but sadly the people I used to follow stopped posting.
We might be transitioning to a world where trust has value and is earned and stored in your reputation. Clickbait is a symptom of people valuing attention over trust. Clickbait spends a percentage of their reputation by trading it for attention.
In a world of many providers, most people have not heard of any particular individual provider. This means they have no reputation to lose, so their choice to act in a reputation losing manner is easy.
Beyond a certain scale when everyone can play that game we end up with the problem that this article describes. The content is easy but vacuous. There are far more people vying for the same number of eyballs now.
The solution is, I believe, earned trust. Curators select items from sources they trust. The ones that do a good job become trusted curators. In a sense HackerNews is a trusted curator. Reddit is one that is losing, or has lost, trust.
AI could probably take on some of the role of that curation. In the future perhaps more so. An AI can scan the sources of an article to see if the sources make the claims that the article says it makes. I doubt it can do so with sufficient accuracy to be useful right now, but I don't think that is too far off.
Perhaps the various fediverse reddit clones had the wrong idea. Maybe they should in a distributed fashion where each point is a subreddit analogue operated each with their own ways of curation, then an upper level curation can make a site of the groups they trust.
This makes a multi level trust mechanism. At each level there are no rules governing behaviour. If you violate the values of a higher layer, they lose trust in you. AI could run its own curation nodes. It might be good at it or it might be terrible, it doesn't really matter. If it is consistently good, it earns trust.
I don't mind there being lots of stuff, if I can still find the good stuff.
I predict a renaissance of meeting people in person.