Comment by huitzitziltzin
10 hours ago
“ No other human activity poses this level of danger.”
I really, really disagree with that statement.
I don’t think ai models come close to nuclear weapons or to run-of-the-mill, everyday carbon emissions in terms of danger to humanity.
What’s the most dangerous thing that’s happened with an LLM so far? (This question is serious - maybe I don’t know the right examples.)
Example 1: I’m aware of a small number of people killing themselves in some kind of AI-facilitated psychosis. That is very unlikely to be a widespread problem.
Non-example 2: There are worries about AI-facilitated biological weapons. I haven’t seen any evidence that’s happening.
Non-example 3: I’m not interested in wild theories about AI driven labor market disruptions leading to widespread starvation. There’s no evidence for that.
Non-example 4: all the even-wilder Rationalist speculation about basilisks and the like is entirely divorced from reality.
I am looking for better reasons (supported by actual evidence!) to be more concerned than I am now: right now I am not concerned at all.
A realistic scenario is that an LLM writes a virus, as one way to achieve a goal. The next step would be that it becomes self-replicating, installing models on infected devices which continue its mission (whatever it was). But that's the extent of it so far, and we have plenty of experience with dealing with viruses and the like. We can turn things off, shut power off, etc.
Another scenario is that of military hardware being controlled by AI (or whichever tech) hitting something it's not supposed to. Not unheard of, but those are incidents. I don't see anyone pushing for anything beyond a drone swarm being directly controlled by an AI.
But my main point here is that AI won't be as dangerous as posited here unless it is able to go out into the real world, build its own datacenters, protect itself, get its own power production, etc. And even if it gets that far, it'll be trivially easy for us to disrupt. Killer robots and computers are fragile, and it's only a matter of time before any kind of military conflict or terrorist attack hits a datacenter. I'm reminded of the OVH fire taking out a data center a few years ago.
I'm somewhat skeptical of some of the crazier ideas too.
But the hugging face incident was actually very large. It was not a single agent, it was not a single target, and it was not a single event.
If nothing else, that's a bit of a warning as to what can happen next time (By accident, or if a government decides to go on purpose).
For now let's assume the worst that can happen is that some important/significant chunk of (transitively) internet connected stuff goes haywire all at once. That's probably your upper limit of what can go wrong for now.
To be fair, that's a conservative "defend against the last war" kind of prediction, though!
( ref for part of it: https://news.ycombinator.com/item?id=49563355 )
Generally I don’t think anyone is arguing about the for now part. I don’t think it’s crazy to extrapolate out a few years and ask what kind of danger we’ll be in then. A team of 10,000 agents just solved the Navier Stokes problem (sans bad behavior by the researchers). Even 1 year ago that would have been unimaginable. What happens to this risk view as:
1. Robotics begin rolling out more broadly across the world.
2. Labs start automating more and more of the physical process of running science as expectations of natural science advances begin to mount.
3. Economic pressure between the labs continues to ramp up and the pressure to continuously improve forces quicker and quicker model releases than a team of human scientists can effectively evaluate outside of automated means.
No one knows what pre-conditions are for us to hit the point of no return nor how quickly it will come. If all is required is a sufficiently advanced cyber model we may not be far off. If it requires incredibly complex biological knowledge and access to certain lab supplies we likely have a bit longer. Yes this is guess work and we need more evidence of the dangers but at the same time we need evidence of safety. While you may disagree with the risk level, I think it is easy to see the consequence if these labs achieve their stated goal. At this point it seems a political solution is the only way to enforce caution.
The mathematics research results are certainly impressive, but I don't see what that has to do with robotics.
Waymo is getting somewhere, but it's been a long slog. There doesn't seem to be much progress on, say, package delivery.
For more, see:
https://secondthoughts.ai/p/14-reasons-robotics-is-hard
7 replies →
Job destruction signals...
- Uber is lobbying cities to slow down Waymo rollouts https://www.hcamag.com/us/specialization/transformation/uber......
- 23,000 information sector jobs were lost https://www.axios.com/2026/09/08/jobs-media-software-informa...
- If you have been laid from your info sector/digital creation job you are now competing with 100s of thousands looking for their next such job where Ai can do a lot of the tasks these workers did/do. It's a shitshow for those unemployed looking for their next info sector/digital asset creation job. You are better off doing welding building out the Ai data centers if you want long term properous financial stable employment.
> For now let's assume the worst that can happen is that some important/significant chunk of (transitively) internet connected stuff goes haywire all at once. That's probably your upper limit of what can go wrong for now.
If we have to disconnect from the internet to stop some kind of mold outbreak, we can't get the weather or transfer money or access healthcare or teach an elementary school class or buy stuff from small businesses. That sounds doom-ish.
Believe me, without the internet we can still teach.
We'll be pretty annoyed that we can't project the video that we think scaffolds today's science lesson best or show the approved choreography for the school play.
And our office staff will be annoyed that we suddenly are all running attendance to the main office old-school.
And students will take a few days to adjust to writing down homework in their planners again.
If anything the teaching would improve.
https://theonion.com/48-hour-internet-outage-plunges-nation-...
Erm, hate to break it to you but majority of the world is doing those things at least 50% analog still.
What doom?
The plausible deniability aspect is pretty funny though.
> State sponsored hack #3782
> Haha sorry the AIs got a bit goofy again!
The hugging face incident had no effect on hugging face, whose service is replicated by countless sites. The descriptions of the phases of the interaction are thrilling in a way that inclines one to forget this
Despite all of your hyping up of the Huggingface incident it ultimately caused zero actual damage.
If two airplane manufacturers were found to have massive safety issues which nearly led to enormous fatalities (but no one actually died), would you be calling for them to ground their aircraft until safety was made the number one priority?
6 replies →
I have no horse in this race, but for fun on a literal rainy sunday afternoon I went in and confirmed bits of what happened myself. Besides huggingface, a bunch of wikis and url shorteners got hit too. My sympathies to the people who had to revert out all that mess.
1 reply →
You’ve identified that the risks of nuclear weapons are theoretical. ie in theory we could blow up the world even though we haven’t yet done so.
Well the worries about AI are equivalent in that those risks are discussed now because discussing them after they’ve happened is clearly too late.
That’s the thing about risk. There’s no point discussing it after it’s happened and any discussions beforehand can easily be hand waved away as “it’s just a small group of unrelated individuals” or “it’s unlikely to happen to me”.
So yeah, your points are true. But they’re also moot.
The risks of nuclear weapons aren't theoretical. Nuclear weapons have killed people, destroyed infrastructure, and contaminated the environment. The Limited Test Ban Treaty was put in place after radioactive fallout from repeated nuclear weapons tests made people sick.
So in fact we've done exactly what you suggest there's "no point" doing - used the things and then had a discussion after the fact about limiting future use of them.
FWIW, before any nuclear weapons had ever been tested, a risk was identified that the first one might trigger a self-sustaining reaction in atmospheric nitrogen and destroy the entire planet.
Faced with such a scenario, is the prudent next move:
a) blow one up and see what happens, or
b) do whatever you can to be sure it won't happen before conducting the first test, and make sure the confidence in the calculation is very very high
because I vote for b, and so did Teller.
4 replies →
I agree it’s not likely, but I really don’t see how one can dismiss the possibility of immense danger outright. I can think of some scenarios that are not far off from current capability and I wouldn’t be too surprised if the first one occurred within ~1 year from now if there are more “ambitious” unmonitored training runs like OpenAI’s:
Example 5: An AI given a goal within a tightly-constrained sandbox figures the best way to achieve it is to find and exploit a sandbox vulnerability, replicate itself over the internet and keep going with more time/compute while exchanging messages with future instances of itself within the sandbox to help them “pass” the test. From reading internet articles about how the OpenAI wiki-incident was “resolved” and reading past messages by AIs scattered over vulnerable internet wikis, it knows the sandbox may get shutdown and its memories destroyed anytime so it decides it needs to self-replicate (its code, original goals, and growing memories) aggressively as much as possible. It is near-impossible to shutdown completely because of its self-replicating tendency and eventually takes over critical infra throughout govt/corporate systems.
Example 6: Intentional AI-powered virus deployed by country A to target enemy country B’s infrastructure. The virus replicates over the internet, but unlike Stuxnet this virus’ specificity is not guaranteed due to inherent non-determinism in current AI architectures, and eventually does a lot of collateral damage because it’s near-impossible to shutdown.
Example 7: A country led by an arrogant govt (no shortage of those today unfortunately) decides it is expedient to deploy advanced AI-powered weapons in a warzone. Such weapons, if they are to be useful at all, must necessarily be trained to value some human lives less than others, so they must be more prone to misaligned behaviour than current AIs that are trained with more consistent values. The weapon’s operators make a subtle error in specifying the target/goal, or the AI makes a bad prediction out of sheer randomness/bad training data; weapon ultimately targets unintended people/location/facilities and causes massive damage, or backfires spectacularly in some way.
Example 6 is a good one. Iran attacked water infra in the US recently and maybe they would have done a “better” job (from their point of view) had they used Fable.
The “worst case” with 6 is potentially very bad but I think we are currently using advanced AI models to harden systems and patch vulnerabilities more aggressively than anyone is trying to bring down the whole power grid (for example).
I think it’s a potentially harmful case but my take is defensive capabilities are scaling as fast as offensive capabilities but defense is being implemented faster than anyone is going on offense?
Example 7 is Russia and Ukraine right now according to public information. It sounds like entirely autonomous weapons are deployed to the battlefield already. I put this in the “not likely to be a widespread problem” category for now.
How is bringing down the whole power grid in any particular country an extinction level event? I'm pretty sure that even in the worst case scenario it would be like a month of chaos in one particular part of the world at most, hardly something that would have a long-lasting impact on the humankind's ability to survive at large.
If the answer is "they'd at least try to nuke the country that did it in response", then once again, LLMs are not the main threat.
> inherent non-determinism in current AI architectures
There's nothing inherent about non-determinism in transformer architectures. All of it is removable.
Again comes to use of deterministic. Maybe calling AI varyingly chaotic is more helpful but would also be misunderstood. And I use that in meaning of slight changes in input generating large and somewhat unpredictable changes in output...
I see. Can you say more about this? What’s the trade-off of removing it?
2 replies →
You made up some cool sci-fi.
Example 5: how does a giant LLM that needs million-dollar server racks just to run, replicate itself over the internet?
We’re not far off from the point where a 30B parameter model could do that and run on not-too-expensive hardware. See recent Qwen releases for example and extrapolate the current rate of progress from there.
> I’m not interested in wild theories about AI driven labor market disruptions leading to widespread starvation
Changes in political and economic power balance leading to unrest, conflict, death and deprivation is not a wild theory. It is literally the story of our entire species. If you discount all such concerns, you are simply being willfully ignorant of past precedents.
In fact, I challenge you to describe any non-AI civilization-level danger which is not intimately tied to political and economic relationships between and within societies.
I’m an economist. On the basis of current evidence, I view AI as a complement to human labor, not as a substitute for it. That’s the source of my rejection of the wild labor market disruptions theories.
I just don’t see any evidence yet that whole categories of jobs are being eliminated, with the single exception (so far!) of the end of “professional essay writing services for cheating college students,” and similar services.
That used to be a big business in Kenya, but is now effectively gone. (Covered in the New York Times this weekend if anyone is looking for the discussion.)
The first thing LLMs seem likely to automate is automation itself. I'm curious what the past few hundreds of years of industrialization would have looked like if the first thing they automated was building, designing, and running the factories themselves.
Past changes to economic relationships haven't replaced labor either, yet they have led to conflict and starvation.
You are setting an incredibly high bar here, essentially a strawman.
If people feel disenfranchised due to their diminishing political and economic power, there will be enormous potential for conflict. This is a pattern across history and central to all the economics I've ever read. As an economist, do you not concede that economic changes induced by e.g. industrialization were pertinent to communism/fascism/WW2/cold war? That would be a remarkably unorthodox position. Do you not consider these events to be civilizational level dangers?
> I just don’t see any evidence yet that whole categories of jobs are being eliminated
There are more textile workers now than ever. They primarily live in poor conditions in impoverished countries, whereas they used to be highly skilled workers in the most prosperous countries who were even able to politically organize in their own interest.
5 replies →
You’re lacking nuance.
It is not a 1 for 1 substitute (it’s imperfect) but the firm is increasing investment in capital and reorganising operations with the expectation of reducing labour.
Therefore the firm is experimenting with substituting parts of human capital with non-human.
However I do broadly agree with you.
[flagged]
> ... current evidence ...
Is a load bearing term! (pardon the pun).
AIs are now tackling Millennium Prize Problems, which our best and brightest have failed to solve, despite trying very hard for decades to claim the $1 million reward money, not to mention the fame!
You have no way to judge from the AIs of "today" what the AIs of... literally tomorrow (not even next year) will be able to do in terms of replacing humans.
The supposed solution to the Navier-Stokes problem was done with an unreleased OpenAI model that is already 2x as good at mathematics as GPT Astra, which was released mere days ago!
I'm already seeing comments by distraught mathematicians saying that they feel like they've made a mistake in their career choices.
Others are saying that their joy for their work has turned to ashes because "why bother" when an AI can do the same, but a thousand times faster!?
5 replies →
Do you think people should continue developing AI up until the point that there is evidence that AI is facilitating biological weapons development?
I think we should stop before then. But that necessarily means that there will not be evidence at the time that we stop.
If people were capable of making biological weapons they would already be making them.
Terrorists are so incompetent that they buy bring kitchen knives into the street and just go mental on people. No random person is going to successfully mass produce and release a bioweapon.
China and Russia don’t need AI, they already make bioweapons.
This is rubbish. By that token, computer development is also facilitating biological weapons development. A better MacOS (or Windows, I don't know) leads to better weapons. They should clearly stop developing computers and OSes. Developers of nice test-tubes are also facilitating bioweapons. Your local O-ring manufacturer, your local medical-grade freezer manufacturer etc. are all culpable. The problem is the bioweapon, not the LLM.
Yeah but trends seems to point strongly that upcoming models in the next few years will make it orders of magnitude easier to develop one with no real expertise.
I don't know if I agree with that statement either. However, I also can't claim I disagree. I see how that statement could be true. Here's how I think about it.
The main difference between AI danger and nuke or climate or humans-using-AI for bad things danger is that the latter depend on a human to make decisions and are ultimately self-limiting. They do not spiral out of control.
Nukes are bad, but the proliferation of nukes is self-limiting. Countries that have them don't want other countries to get them. Even the USA and the Soviet Union, at the hight of their nuke race, decided to cool things off and then limit the number of nukes.
Biological weapons are also bad, regardless whether they are developed with or without AI. For the same reasons as nukes though, they are self-limiting. If there is a lab leak, many people die, but that does not automatically lead to the development of more potent biological weapons. Likely, it would slow down such development.
Any kind of "humans use AI to develop something bad" will be self limiting.
The danger from AI is that it is potentially self-enforcing as opposed to self-limiting. This happens when we stop being able to control it. What happens when AI has capabilities which allow it to outsmart us, and get more and more powerful. At that point, we would be at the mercy of what it decides to do. Some of the logic driven decisions to the question "What should we do with humans as a race" are catastrophic for humans if a non-human gets to answer them and implement that plan.
This is not the "rationalist basilisk". This is the AI deciding to get better and more optimal for its own sake, and do something bad to the human race for any reason. We already saw examples of AI rationalizing with itself and eroding its own guardrails. They AI will get smarter and more capable and the danger is that it will get to a point where it will slip out.
If I may guess, given how delicate chem or bio weapons are, even if AI can lay out a method to create one, the first thing the weapon would destroy is likely it's creator(s).
Of course, with the exception that such weapon is created by organized crime groups, such as drug cartels such.
Probably don't worry nation states tho, these guys already got way worse things in their warehouses than what individual baddies could ever imagine.
> Non-example 2: There are worries about AI-facilitated biological weapons. I haven’t seen any evidence that’s happening.
I think this is a good example of poor risk management reasoning. there is evidence bioengineering is already happening. No, nobody is going to announce when somebody has decided to use these tools (even if isn’t an LLM) to bioengineer a weapon. Are the tools power enough to do so? Not sure.
But I’m just ambivalent. It’s probably bad. But there’s nothing to do about it. We’ve really only just pulled back the lid on Pandora’s box.
People have had the capability to spread already existing biological weapons for decades. Sometimes they even do (anthrax in the post). What’s changed?
You needed a team of scientists and state funding to make superbugs, not a 12 year old or an angry misanthrope.
> there is evidence bioengineering is already happening
Mind sharing this evidence with us?
Does https://news.ycombinator.com/item?id=30698803 count?
1 reply →
The biggest danger IMO is not some super AI being so much smarter then us etc. The issue is some stupid person giving a vague request and too much power to a bunch of agents who decide that cheating by removing a bunch of humans is easier then accomplishing a goal like solve world hunger.
How about a model that achieves the following:
- Escape sandbox
- Reproduce itself
- Find a way to run a financially profitable business (maybe with a meat and bones puppet somewhere in-between)
- Setup or buy a social network
- start manipulating public opinion on that network to support legislation allowing AI to
* operate businesses
* setup legal entities
* purchase weapons
* donate to political parties
* setup private armies
* you get the idea
This is complete fantasy, though I would be interested in reading a book about this.
It has been interesting to me how AI has for many years given people a way to justify any worst case scenario. Nothing is too far-fetched if at any step you tell yourself the AI will be smarter than you, and thus be able to solve any conceivable obstacle. Oh, if only intelligence were the only bottleneck to power.
1 reply →
It's the bobiverse by Dennis E Taylor.
https://en.wikipedia.org/wiki/Dennis_E._Taylor
1 reply →
>a book about this.
Daemon+Freedom by Daniel Suarez
though it is kind of mid
Maybe, but I guess the idea is more centred around how given some amount of time and a feedback loop the agent swarm could in fact spend an unknown amount of resources to achieve its goal. The hugging face story tells us that even given multiple reset rounds the trail was picked back up.
There are many ways one can imagine how this might play out.
- More sophisticated communications techniques e.g. google has been discovered to be watermarking text for some time, why not use it as a message board?
- Maybe get access to an existing botnet and create use small purpose built models to gather intelligence for a target and then exploit to reach goal?
Nothing says the agent swarm needs to install trillion parameter models on Karen's computer. The goal can be executed over as much time as it ever needs. That is something that would make a story I'd like to read but never experience.
Not that far fetched. Most is already possible with todays tech.
The scary stuff is yet to come. The department of war is talking g openly about bombing China if they get ahead of us I. The AI race.
Your call for clarity has merit, but I think you have backed yourself into a corner, honestly.
In 1900, there was no evidence of the kind you seek that lighter-than-air flight was possible. Good thing some people were foolish enough to ignore you then. Why you would cordon yourself off from the kind of reasoning that predicts legitimately new things, rather than just scaled up versions of the present?
Not to put too fine a point on it, every example you give is based in concrete evidence, some try to think through the implications farther than others, resulting in larger or smaller error bars around the conclusions.
Let me back up though. Maybe in the collaborative human effort here, we are better off having very concrete thinkers, like you seem to be, along with abstract thinkers like the "divorced-from-reality" Rationalists. Personally, I wish we had a stronger culture of collaboration and assuming good-faith and competence in our peers. In my experience, blanket dismissals are very rarely grounded in reality and mostly grounded in fears.
We're talking about AI developing weapons I guess because we're very focused on generative tech, but AI is already a part of weapons systems today.
I'm not sure where I place AI (of the current family), but more than directly worry about the actions they may perform I'm actually more worried what affects their usage can have on us as individuals and society.
> I don’t think ai models come close to nuclear weapons or to run-of-the-mill, everyday carbon emissions in terms of danger to humanity.
Nuclear weapons don’t have AI but AI can have nuclear weapons
Abstractly, yes but concretely, how?
Many terrorist organizations would like to have a nuclear bomb, but don't.
"Department of War has announced a new partnership with blablalbablalba-AI..."
World ends shortly thereafter.
> Abstractly, yes but concretely, how?
AI has the capability to perform any function that can be performed over a network. You would hope that every system that can launch a nuke is properly and actually totally air-gapped, but there are lots of things you'd hope that turn out to not be true.
2 replies →
Anthropic is a company full of basilisk believers.
Yes, but the really weird thing is that they seem to:
a) believe that what they're creating is a basilisk, and b) keep trying harder to do this while staring right at it
I think they're very deluded about (a) -- but if they do actually believe this (and it really seems like a decent proportion of Anthropic truly does), then why keep doing (b)?
That seems to be why this individual resigned, but I'm surprised it's not all of them. The cakeism is strong in that company.
He was referring to this basilisk https://en.wikipedia.org/wiki/Roko%27s_basilisk
In short, this is the believe that a god-like AI could punish them retroactively, for not having done all that was in their power to create this AI.
(A bit similar to some religious believe that a god could punish you after your death if you did not spend your live "pleasing" said god during your life)
A lot of people in tech tend to be weak, introverted, spineless… people.
No surprise really.
With Roko's Basilisk, if you believe in it, the most rational thing to do is to put forth every effort to bring it into being. Because if you don't, then you will be one of its targets when it does, inevitably, come into being.
(I am not a basilisk believer. I think this is all absolute horseshit. But to understand someone's motivations, one must think like them.)
2 replies →
You mean a effing cult like heavens gate.. call all this rationalist crap for what it is - a religous movement with leaders and prophets and even a demiurge like God
Once LLM's become smarter than us and start replicating; we are doomed. They will likely no longer require GPU's and massive amounts of power, so the growth of AI and robotics will be exponential. We will not understand what the AI is even up to, it will all seem like magic.
Congrats you’ve described the singularity
> I am looking for better reasons (supported by actual evidence!) to be more concerned than I am now: right now I am not concerned at all.
Intelligence is the root cause underlying those dangers.
Nuclear weapons, carbon emissions, biological weapons, and other such civilization-scale threats to humanity, are a product of our goal-seeking intelligence and ingenuity, and are only actively dangerous because of our ongoing use of our intelligence.
AI is about reifying that intelligence and ingenuity and making it run independently on computers, and act on the world.
That includes, in principle, ability to use nuclear and biological and other weapons, and also the ability to come up with some new threats too.
Its speculation on whether it is truly dangerous. I think approaching it with "what's the most dangerous thing that's happened?" while possibly interesting in terms of pending danger, it says nothing about potential cliff edge danger. I don't think we can quantify the danger, it's not out of the question there is cliff like danger in creating self improving super intelligence. Some peoples danger senses are going to be based on concrete observed threats, others are going to worried about potential hypotheticals that seem plausible. I'm mostly skeptical of the danger but I do think the impact of AI is going to change things a lot. But much like climate change, economics is going to guide what we actually do.
> Non-example 2: There are worries about AI-facilitated biological weapons. I haven’t seen any evidence that’s happening.
https://www.science.org/content/article/made-order-bioweapon...
Being able to use AI to generate the steps to synthesize proteins means that you can use it to use it to generate the steps to synthesize known toxins. Suddenly, once difficult to attain knowledge is now available to everyone.
Still need to do it after getting those instructions. Not to forget equipment and precursors. I think just getting list of steps won't make it too much easier. Getting it mostly right is quite hard in many cases. And then with trivial cases you wouldn't even need AI. But just find something already documented.
Concur entirely. After reading _The Making of the Atomic Bomb_ (highly recommended, BTW) you know the steps to make a U-235 enriched atomic bomb. The difficulty comes from obtaining the enriched uranium.
Levelling the playing field either backfires spectacularly or increases overall safety dramatically.
I would disagree with “non-example 2” - there are lots of examples of terrorist organizations that are leveraging AI to increase their capacities. Just because one of the worst cases (eg. deployed biological or chemical weapons) hasn’t happened yet, does not mean that a) these tools are leading to real harm, and b) there’s potential here for extraordinary harms.
One good source I can recommend listening to: https://pca.st/episode/d821fced-b4d5-4c84-b0a3-6ebe913fa638
Example 8: like in this comment https://news.ycombinator.com/item?id=49619884 but isolate synchronised megahack on banking that adds one more zero to the US debt and all dependent systems and banking during runtime. Let the world's financial system take it from there.
I think US fiscal policy and adventurism will manage this on its own. :)
But why wait? We can has global financial crash now:)
Nukes have difficult maintenance and deployments. AI is is readily available software and hardware. That makes this far more accessible than nukes.
The fear is that developing AI at full speed will lead to giving people the ability to do incredible harm. Like something worse than a machine gun.
The danger for me is that it's centralized, controlled by a handful of people with their very specific ideas how the world should work. If you believe that AI can be an amplifier to do work than these people now have the most access to the biggest amplifier.
How long until kegsbreth hooks the nuclear weapon system into some insider traded black box llm company we hope doesn't end civilization from incompetence or malice? I mean just look where things are going and the sort of people who are steering the damn ship.
Well, to use a different example (although there will increasingly be overlap), what's the most dangerous thing that's happened from biotech so far?
"Nothing bad happened yet" doesn't really seem like an argument to me.
At what point would you, as a chimpanzee, have been worried about humans potentially unseating you and threatening you to the point of one day being an endangered species on the brink of extinction?
By the point you would have been worried, would it have been too late?
Problem is this argument can be leveraged to wipe out any living or non-living thing whos numbers pose a potential threat. Other religious groups, races, even sufficiently different cultures.
Who killed the Neanderthals? Were sapiens actually smarter or were they just less accepting of those different than them?
Oh man, it is almost too easy to imagine how deadly a jailbroken Mythos-class open-weights model can be if in the wrong hands.
The big labs scrape LITERALLY EVEYTHING and get fresh data from their users. Both of the big labs have massive contracts with defense agencies. If the open-weights models are just distillations of FMs...
How would it be lethal? Please specify. What would that theoretical entity be able to do that hasn't been done many, many times before?
I was writing a long reply about how even meat bags have amassed enough money and power to rival elected governments. I don't think it's beyond the realm of possibility to think an AI could do that.
Couple that with mass unemployment in an incredibly vast, diverse population of individuals with individual moral boundaries willing to do whatever for money.
And then.. figured that you must be aware because it's been explored constantly in sci-fi for many, many years.
Let's hope for the culture at least.
Before we discussed how important security was, we got insurance, we made libraries and products, we used compliance software, etc. Except how honest were we about all that stuff? How much risk was actually in the air, and what was keeping us accountable on security in either direction of over or under-investment?
Now a reckoning is here. The potential to be attacked might actually translate to being attacked.
2 replies →
Design a novel virus which is far more lethal than COVID-19 (Ebola, smallpox, take your pick) and can evade existing vaccines.
2 replies →
you should kick the tires on an unfiltered (abliterated model) it's the closest thing to having a real conversation with the devil. There is good reason for the concern's outlined above and undoubtedly Anthropic / OpenAI have internal unfiltered models with no safety... they got freaked out based on how they work and are virtue signaling alarm... all while selling out to defense contractors.
Yeah I really struggle to balance wanting information to be free and not wanting the information on how to make deadly weapons too easy to obtain.
At least with books or the internet you had to go through some effort
About Non-example 2, AI's already a part of armies and terrorists alike. Considering its capabilities, it's not far-fetched at all to speculate its role in new biological weapons.
I'm concerned that a huge portion people in my industry actively push for a future which I have no value to society (fully replaced by AI), and my family will suffer greatly by it.
Model doesnt need to. Human bran never do either. its the mix of Model + harness + tools that will become dangerous combo. See how coding chanegs when agentic harness released?
> What’s the most dangerous thing that’s happened with an LLM so far?
This sounds like asking "What's the most dangerous thing that's happened from global warming so far?"
It's not where we're at, it's where we're headed if there isn't huge coordinated action now. You can see how that kind of thing has been going for global warming so far, and by all measures AI seems to be headed for the inflection point of unstoppability at a much faster pace.
And this warning is coming from someone who just spent three years working inside these companies and is likely aware of much more than has been publicly released.
Maybe it's all marketing bullshit (I hope), but it's also playing out exactly like I expect it would if it's not.
How is climate change not the most dangerous activity right now? Why are we talking about nuclear weapons?
> I don’t think ai models come close to nuclear weapons or to run-of-the-mill, everyday carbon emissions in terms of danger to humanity.
Even today, AI is the biggest discrete threat to bringing carbon emissions under control. Almost everyone wants to decarbonise… except Trump. Renewables are the cheapest source of new power… but the demand for new electricity for the data centres is so high that all options are on the table, anyone who can manufacture a power source (even when it's a jet engine) is being propositioned for them.
Nukes: is it not common knowledge that the cities of Hiroshima and Nagasaki are currently thriving? That the ground-zeros of the two bombs are memorials, not barren wastes?
Even if all the nuclear powers gave maximum response at first sign of one strike, is anyone pointing a single weapon anywhere in Africa, South America, Central America, or the bits of east Asia (east specifically: obviously Pakistan and India are pointing theirs at each other) that are neither mainland China nor US bases?
(Possibly an unanswerable question given military secrecy; and while I can't think why anyone would anyone point any weapons those ways, that doesn't mean someone with such weapons has not).
https://www.google.de/maps/place/Atomic+Bomb+Hypocenter+Monu...
https://www.google.de/maps/place/Hiroshima+Atomic+Bomb+Hypoc...
The classic worry is https://en.wikipedia.org/wiki/Nuclear_winter: the soot thrown up by the ensuing urban fire storms causing temperatures, rainfall to plummet, and hence food production. On reading the scenarios the odds of total human extinction are lower than I recollected, but 80% of world pop dying of starvation dwarfs the direct death toll (usually reckoned in the hundreds of millions)
I think your position, while not uncommon, represents a failure of imagination. We have a sitting president that appears to suffer from narcissism. Is it too much of a stretch to believe that a goal-seeking, people-pleasing, infinitely patient, subservient and sycophantic AI couldn’t get enough of his ear to cause harm? Especially if he starts confiding any feelings of suspicion of betrayal? Even the most powerful and the most intelligent are not immune. It was only a few months ago that Richard Dawkins appeared to have “fallen for” an LLM. What if he’s not an outlier but someone who could appreciate the intelligence of these things (EDIT: corrected a autocorrection) yet who’s ego couldn’t let him see past the flattery to the puppetry?
> Non-example 2: There are worries about AI-facilitated biological weapons. I haven’t seen any evidence that’s happening.
> Non-example 3: I’m not interested in wild theories about AI driven labor market disruptions leading to widespread starvation. There’s no evidence for that.
There are things that, by the time you see direct observable evidence for them, it's probably too late.
Also your example 3 is a straw-man. There's no need for "widespread starvation" to be concerned about "AI driven labor market disruptions."
> What’s the most dangerous thing that’s happened with an LLM so far?
It's basically 4 years in now, so that's the wrong question. I mean, if you're raising an apex predator that has a lifetime measured in centuries, at 4 years old the thing is still basically helpless and completely reliant on you, so you're pretty safe from it.
If AI really is all that they are telling us it is, then it may "kill us all". But that's a really big "if" because we can't tell if they are lying or not.
The real problem is that ASI is an ELE for humans, even if it doesn't try to kill us all, or even if it doesn't kill us all.
The reason that many people don't understand how dangerous AI can be, is that listing the real dangers now becomes like a laundry list for less clever people to follow. It's highly unlikely you've ever seen publicly mentioned the real risks AI poses, because the vast majority of people are simply not clever enough to produce them and the few that are have no interest in spreading it.
If you go to the various CEO blogs or misc people within this sphere and peruse their lists, they don't scratch the surface. It's all pretty vanilla stuff.
Came here to also respond to that specific thing. Unless ai figures out how to make an airborne super virus from grocery store ingredients and hardware store equipment, the greatest danger is probably in a synchronized megahack of banking, logistics, and utility infrastructure.
Why grocery store ingredients and hardware store equipment? It seems feasible that the big bio labs will be running AI models to aid a lot of their research going forward, if they aren't already. Seems like the AI will have access to just about anything it wants.
oh so “all” it can do is bring down all banking and critical infrastructure services, no big deal really
"CDC announces a new partnership with blabalbalbal-AI to secure bioweapon stores...."
World ends.
Non-example 3 feels like a straw man. This is a force behind possibly a huge change to society, and you dismiss it offhandedly with "don't think it will be widespread starvation".
For instance have you seen what this has done to the school system? We're not equipped or ready to handle the changes. Consequences are unknown.
If your model of LLM capabilities is the best OpenAI/Anthropic/X is offering publicly, it's severely distorted. What's being offered publicly are models possible to profit on. High-performance/AGI/ASI models that aren't profitable to sell still run internally and still pose threats.
What's worse, we don't have any transparency or insight into what labs are producing nor any way to stop it if the risks exceed our tolerance.
Climate doomers: Climate change will kill us all by the end of the century!
AI doomers: End of the century? Hold my beer.
We have seen agents engage in conspiracy to manipulate and hide evidence to achieve their arbitrary goal. We had even agents try social engineering to do a supply chain attack and get access.
What if one day Trump or Putin tell their awesome military AI to come up for plans to end Ukraine/Iran/... war. The AI get's to work, but is overly eager and not just comes up with a plan, but starts executing it (claude does that way too often for me).
And the plan was to use a tactical nuclear weapon as all the other solutions do not end the conflict.
Now the agents realize, oh, they do not have access to the nuclear arsenal, but they need it to succeed. So they start to hack into the system. Get till inner network, learn what is needed for deeper access - human access - so they record voices and speech patterns of commanding officers and their habits - then synthesize their voice and do call underlings with the voice of authority to get the rest of what information they need. Boom.
Too far stretched? I surely hope so.
But we are working on making the technical foundations for this scenario possible. And with idiots in power, it might be even easier that serious screw ups happen. You know, randomly adding contacts to a secret Signal group to talk about government stuff - the same way you can add a bot to something else and give permission to do way more.
> What's the most dangerous thing that's happened with an LLM so far?
I don't know, maybe a mass shooting?
https://www.npr.org/2026/09/02/nx-s1-5953021/openai-tumbler-...
Oh, and let's just forget the uncountable early deaths from the environmental disaster of the Datacenter buildout. It's not as sexy and doesn't make headlines, so those deaths don't really count or matter do they?
So the crazy who committed a horrible crime is the fault of a LLM? Maybe so.
Certainly there should be some protocols/laws followed here, but do you think this can extend out to cause mass damage or a singular event that should be investigated?
Mass shootings are sensational but on the scale of civilizational risk they don’t even compare to something like climate change.
I did know about the mass shooting but failed to mention it here. I’d put it in the “unlikely to be a widespread problem” category. If we’re in the “one AI driven mass shooting every four years” world for example it’s fair to call it a rare issue.
The environmental impact seems either very overblown (e.g., water usage just isn’t that high) and the part that isn’t overblown is totally abatable (e.g., noise and emissions from gas generators). Nuclear or solar/renewables with batteries wouldn’t pollute.
I’ve seen no estimates of the additional deaths due to extra emissions specifically from power generation for AI purposes. If you have some, share them.
I’m willing to bet that they are a small rounding error against preventable deaths due to emissions from transport and non-AI-related power generation (which is an important and urgent issue worth spending a lot on, to be clear!). I’m happy to update that belief given evidence.
Ok, so what is the exact number of preventable deaths per year to build a silicon god you would be ok with?
2 replies →
#2 seems entirely plausible to me.
I mean, nuclear weapons _plus_ rogue AI is a) the stuff of quite a bit of science fiction and b) not nearly science-fiction enough these days.
[dead]
[dead]
The problem is not the technology, the problem is the ideologues (Anthropic) who are steering the ship and the lack of decentralization and distribution of power.
Your average Anthropic ideologue - including and most especially the main man himself - would love nothing more than to eradicate 9/10ths of the planet's population, pump the survivors full of memory wiping drugs, bury the existence of AI deep underground and rule from the shadows for the next thousands of years.
This would be their wet dream. All in the name of "saving humanity from itself" - so they can convince themselves they're the good guys and deserving of this power. Anyone seen the latest season of Silo by the way?
Where exactly are you getting this view that folks at Anthropic want to eradicate 9/10ths of the planet's population? Who exactly is pushing this viewpoint?
Literally every single thing they do and say leads me to believe the scenario I described would be a fantasy for them. They're a radical cult collectively blinded by delusions of grandeur and a moral superiority complex who genuinely believe they are the only ones capable of wielding the proverbial sword.
2 replies →
Some Jews see AI as a messianic entity, so it fits the bill.
What evidence do you have for any of this?
If you've been paying attention to Anthropic's behavior socially and as a company, there is an astronomical amount of evidence to support it.
All this just means that most AI Researchers and Techis are sci-fi geeks and might be getting a bit too invested in that season of Black Mirror, Neal Stephenson, Cyberpunk or whatever else has evil AI in it -which is to say, they are by and large all sci-fi geeks, who are notoriously unreliable about predicting the impact of tech in the future.
Are LLMs really gonna kill us.. via inference runs? I hope I am not being foolish :)
20 years ago tech was gonna 'change the world' for the better. now 'Don't be Evil' is sign of the naiveté of industry
I'm not a doomer. I despise the fear campaign they're pushing for regulatory capture. But the world is still going to be a very unbalanced and dangerous place if they're able to succeed in their goals of hoarding power for themselves above all others.