It won’t kill us as a species, but its use may lead to disasters that were unintended and didn’t need to happen. Much of the regulation in other industries is due to that kind of thing happening. Worse, the companies making them are trying to get regulation to curtail other parties development for economic advantage rather than regulation on use which will steer the entire industry.

disasters that were unintended and didn’t need to happen.

As you say, every “scientifically based” industrial innovation / breakthrough has had the potential to cause new kinds of mayhem, and most of them did to some degree or another. Burning rivers and CO2 overload in the atmosphere spring to mind, but industrial “accidents” are part and parcel of industrial activity, always have been.

LLMs are not intelligent. They lack morals, knowledge, and intent. The only remote risk of extermination is if a company with absolute impunity and endless resources told an LLM to hack the nuclear weapons and if knowledge of how to do that was already published online and processed as training data.

LLMs cannot do anything new, they can only regurgitate information thats already been stolen.

You’re confusing morals, knowledge, intent, and intelligence with capability. An atomic bomb has no will of its own either.

idk, every atom bomb I’ve met has one purpose…

An atom bomb has no will of its own either.

That is simply counter-factual. LLMs can absolutely do things that have not been published before, and they can access knowledge that is not in their training data through RAG. Intelligence is debatable, but reasoning models can reliably imitate “thinking” to a degree, and recursive self-improvement might lead to an autonomous advancement of intelligence. That LLMs allegedly “only regurgitate information” is simply wrong.

Or if the military put an LLM in charge of nuclear weapons. But surely nobody would be stupid enough to do that!

“And that’s why today I’m signing into law the Trump Nuclear Act. Some say it’s the greatest law of all time. Very smart people, even smarter AI. Elon tells me this will save billions, and I’m going to increase my dividend, let’s call it the ‘Nucular Peace Dividend’ to honor Bush, I don’t remember which one used to say ‘Nucular’, but I always liked that guy, and it made me think up how to say Chayyyyna.

CNN get out!”

The only remote risk of extermination is if a company with absolute impunity and endless resources told an LLM to hack the nuclear weapons

Absolutely a huge risk.

and if knowledge of how to do that was already published online and processed as training data.

extremely naive to not understand existing capabilities of 100s of probings and planning based on probings, and multiple strategies, in a swarm working towards a goal.

Still to your first point, the organizations with the budget and impunity, seeking the most powerful tools, are then the ones who will use them for evil, at expense of others, power maximization.

can be so unreliable, it might kill us by incompetence.

Same can be said of Homer Simpson manning the console at a nuclear power plant.

AI is a tool, irresponsible use of a tool is… irresponsible. Just like people puttting “not-self-driving” cars into “supervised driving assistance mode” and taking a nap at the wheel at 80mph.