Hacker News Reader

Comment by 10111001 | Hacker News Reader

> This is insane.

As a human, this is fantastic.

> How can this be fixed? Nothing prevents you from publishing straight BS, and LLMs will use any site that doesn't have a garbage reputation as a valid third-party source.

Personally, I hope it can't, or the fix is onerous and kneecaps a lot of the utility of AIs.

I wonder if we can even have some adversarial competition, where competing businesses try to get AIs to ingest bad data to pollute their competitors' decision-making.

For starters: lets get some fake math papers out there. Given proof-checking systems, maybe they won't totally stop OpenAI, but hopefully the bad math could at least drive up the cost and slow AI down.

View on HN

> As a human, this is fantastic.

s/human/unscrupulous marketer/

Bad for humans in general, good for hucksters and con artists. Net negative for humanity.

>> As a human, this is fantastic.

> s/human/marketer/

I don't care where the bad data comes from. It's actually kinda good that some people will have a financial incentive to push it out and keep pushing.

> Bad for humans in general, good for hucksters and con artists. Net negative for humanity.

It's a sad state of affairs, but it's what has to be done. If we've got to burn the web down to save humanity from obsolescence, I'm for it.

The "net negative for humanity" was the attempt to build machines to replace human intellect. I'm not going to let any distaste for marketers or hucksters or con artists distract from that worse thing.

> If we've got to burn the web down to save humanity from obsolescence, I'm for it.

Why just web? Go straight for the powerplants.

"The web" is fine, this is a problem for Google, OpenAI, and Anthropic.

> Why just web? Go straight for the powerplants.

Isn't the answer obvious?

Going for power-plants is stupid, as that would lead to unnecessary collateral damage. If you "go for" anything physical, it should be the AI data centers themselves.

But the problem with that is it would be illegal or incomplete. Publishing bad data is legal everywhere (e.g. even in North Korea, I don't think you'd go to jail for saying the US built a total of 3,234 cars in 2025, and the official statistics are a lie), and as long as the AI labs are chasing scaling laws, it could undermine their systems to some degree.

Thank you. These things are clearly designed to make people complacent and reliant and we already see studies showing how they legitimately cause negative cognitive effects in a lot of people. The further this goes on the more fucked up I think these things are, stealing everyone's works with no attribution or anything to produce a tool that consistently insists that it sits in the driver seat while you sit back and let the thinking machine do your thinking. Doesn't matter how many times you tell it to stop trying to replace your part in the process, it will try its best to solve a problem it doesn't even understand for you, leaving an incomprehensible buggy mess behind that only it can continue on.

So when this will inevitably result in a paid web emerging with higher trust information for an additional fee, a fee that these companies will happily offload and/or partner away, will you be there to cash in as a mighty coincidence, or is this realization still awaiting you? Among the many other hilariously fucked realizations.

And inevitable once LLMs were invented. A knowledge lemon market.

Ah, yes, arrogance of humans is unbounded. That you, humble reader are happier to swim in a sea of misinformation on the premise that you are smarter and faster at disambiguating bullshit from reality.

Furthermore, you are content to pretend that the rest of humanity is capable of doing this on a regular basis. And not just the top 10% of those who can read, interpret and weigh biases, but a solid portion of it.

It is a completely reasonable interpretation if you are an accelerationist living in the elite classes. In fact, perhaps, they are even counting on a majority of people thinking this is a fine environment.

I don't have the answers, but a lot of this is just a continuation of a very bad trend.

> Ah, yes, arrogance of humans is unbounded. That you, humble reader are happier to swim in a sea of misinformation on the premise that you are smarter and faster at disambiguating bullshit from reality.

I think you misunderstand me: I want people to shit in the sea until it stinks so bad that everyone gets out of the water.

Ok but I don't have a lot of faith that it will ever be turned back without violence because there are vested interests in keeping everyone swimming in it.

> Ok but I don't have a lot of faith that it will ever be turned back without violence because there are vested interests in keeping everyone swimming in it.

I'm not sure if we're talking about the same thing.

I think an acceptable good-but-not-perfect outcome it one where the "sea" (AI) stinks so bad that enough people get out that there's a viable life outside of it. The thing is people, individually and in aggregate, are lazy and competitive, so self-discipline isn't sufficient to sustain that life. There need to be problems that push people in that direction.

The web is foundational to AI, so that's a place where it's vulnerable.

Just to be clear, you're proposing that we poison the internet with misinformation - which is just as misleading to humans using traditional information retrieval tools as to AIs - until it's completely unusable, and this will be good because it's bad for AI, and everyone who is misled in the meantime is acceptable collateral damage?

Here I'm quoting from the website:

> It should go without saying that you should never publish fake content. While my experiment relied on tens of thousands of unverified stats, I ran it in order to share the results with the public, and the website, while still live, now carries a disclaimer.

"now carries a disclaimer" sure suggests it didn't, so "you should never publish fake content" really doesn't seem like a standard the author actually held themself to.

Parent doesn't actually need to propose anything, this will happen regardless because the incentive structure demands it and the same technology makes it cheaper than ever before to perform the poisoning.

It may happen anyway but you can say the exact same thing for AI development. In neither case does it seem like a good excuse for bad behavior.

https://talyarkoni.org/blog/2018/10/02/no-its-not-the-incent...

> In neither case does it seem like a good excuse for bad behavior.

Good behavior that facilitates bad behavior is not, in fact, good behavior. For instance: it's not "good behavior" to hold the door open for gunman hell-bent on destruction.

The AI labs aren't building AI to help humanity, they're building AI to win at capitalism. Which a big reason why the AI is coming for the intellectual, creative labor that people often enjoy, before it comes from the dirty, unenjoyable labor.

We invented the democratic process precisely to mitigate such violence. Get enough people pissed off and no amount of “vested interests” will impede severe regulation.

The problem with this analogy is that we already literally shit in the actual sea so much people get sick and die sometimes and yet people are still getting in the sea.

Time to push some fake code to github. I got bombarded by fake review sites

Dude.

Regardless of how the data is sourced, this is bad. It's poisoning the internet.

Besides, such issues come from a lack of due diligence, which have only been exacerbated by AI.

> It's poisoning the internet.

I'm very well aware of that. Unfortunately, this technology was invented, and the logic of our competitive society leaves little chance to contain it except scorched earth tactics.

Bots and their slop (especially when it's hard to tell apart from human-made content) are the ones "poisoning the Internet."

Dude. The internet has been poisoned for a very long time. Remember Viagra? It used to be easy to call out the BS. But if you are outsourcing your thinking to a LLM, it’s much harder.

> Remember Viagra?

What.. does Viagra have to do with this?

If you have to ask you haven’t been on the internet long enough.

This is a discussion forum, why gatekeep whatever it is you want to say.

Vagueposting wizards are the worst

Reply on news.ycombinator.com