Technology

I was in machine learning for a bit, up until venture capitalists eviscerated research teams chasing ways to make money from it. While there's definitely a degree of overhyping risks for investment, there are absolutely some real, tangible risks. But the dangers aren't like skynet/terminator, its more of a slow decline in knowledge combined with an increase of predictive surveillance (https://arxiv.org/abs/1902.10739 as an example of why predictive surveillance should worry everyone)

Every mistake and error in token generation is getting confounded with human text, and being fed back as part of the training loop. Combined with the sheer amount of botting of the modern internet, you get a feedback loop of error being used to train, followed by output of that error dominating the next training cycle as bots drown out voices of the knowledgable. For some basic things it's easy to see the error, but in any expert domain the ratio of bots to experts is large enough that expert knowledge won't stay within the training distribution.

AI industry makes money from doomsday prophecy about it being so intelligent that it will create a point of no return: the [singularity]. It is a magaian claim, a claim that is so increadibly stupid that one does not know where to even begin; and if you do begin, you've already lost, because you have given attention to the claim. Lets not give these stupid fucks attention.

AI is not dangerous because of the [singularity]. It is dangerous because of its power consumption, automation of war, turning everybody to unthinking, surveillance automation, burning books, the economic instability it will cause due to the bubble bursting, water consumption, destroying the internet, it gaining responsibilities it is not qualified to take and so on.

But we can see how the story of the [singularity] and similar stories helps AI companies, because now, instead of criticizing real unwellnesses, we criticize a fictional unwellness that generates discussion regardless of our positions. It is a conversational black hole.

lets get back to the real world

The more shit that leaks out of anthropic the clearer it is that place has become a cult

How will "AI kill us all"? I can see it creating harm and getting people out of jobs. But kill us??

Everytime I see an article about the dangers of AI, I'm thinking of some version of this comic every single time.

And the reason it is dangerous is generally because someone is convinced that Ai models are not as shown above.

AI isn't a danger to humanity, per se. Humans abandoning their duties of being responsible stewards of systems is a genuine source of danger though.

We already have algorithmic abuse where people are subjugated by welfare and debt systems with little oversight or flexibility. Some organisations even breaking the law and going against their stated missions to achieve goals and maintain order.

Social murder is the concept that society can kill a person through stress, pressure, abandonment and indifference to the individuals struggles and needs. We're seeing a lot of it today with the declining health of younger people being especially alarming.

AI can scale the two things I've mentioned beyond limits set by current bureaucratic systems, I.e. the slowness of people running a system. Systems which already kill and harm people without AI. The small erroneous tendencies of logic and algorithms in systems will be scaled up, further and faster. AI can enable algorithmic abuse and social murder on scales comparable to mass murder.

They want regulation to create a cartel. Failing that they want hysteria to increase pre IPO valuations.

As Patrick Boyle said. It would be like Coca Cola, Pepsi and Irn Bru getting together to beg for regulations because their soft drinks are just too delicious.

If someone stands to benefit financially from the idea their AI is so powerful that is scary, they're probably going to bullshit about how powerful it is

So apparently we're using up all our resources in order to create something that might kill us all.

Sounds like a fantastic idea.

I'm going to go cry now.

They are lying about their fear. That's all they can do. Stop believing any claim of theirs. Stop being manipulated.

They are snake oil salesman. They sell vaporware. Don't fucking buy it.

It's good to see Doomsday Cults back in vogue. Too bad Robert Anton Wilson is not around.

I've heard a few explanations for the "AI will kill us all" narrative.

1. AI techies are in a cult. They genuinely believe this stuff because they're crazy.

2. It makes the AI sound powerful and impressive, which is ultimately good for the AI companies.

3. They're starting to face competition from open-weight models, so they want to scare politicians into crafting regulations. They think that they will be able to influence the regulations to favor them and squeeze out their competition.

And, of course, there's

4. It keeps people talking about AI. Because whatever you do, don't quit talking about AI, because if you quit talking about it, the line might go down! And the line going down is very, very bad.

No matter how often I read this kind of stuff, I simply don't buy the 'AI is dangerous for humanity!! We gotta stop their development or at least press for sane guardrails!!!' narrative.

it's pretty disrupting tech, not without risk. general population is behind on the real capabilities of AI and tech companies are doing smart marketing which gives a big spread in perception.