Jensen Huang: Those warning on AI apocalypse "have ulterior reasons"

Are you denying that the weapon could have destroyed humanity, if built and detonated?

> Cigarettes, junk food, social media, fossil fuels, nuclear reactors and cults all have a higher ROI than OpenAI or Anthropic.

You're repeating "ROI". This suggests to me you think it's shaped like an argument. I do not understand why. Of course they do, OpenAI or Anthropic are in a rapid growth phase, most startups take 5 years or so to become positive ROI and even though OpenAI is theoretically older it spent most of that as a not-even-slightly-for-profit.

Also, nuclear reactors are heavily state subsidised (if you include guaranteed price for energy delivery); they're basically there to give the hosting nation experience handling materials and a source of neutrons to rapidly make more, so that in the event they find they've got short notice to make a weapon they can, while the energy output is a way to reduce the cost of having the capability.

All of them, including reactors: the point is nobody benefits from being painted as bad. This is not a thing. The sales pitch for e.g. SMRs is not "Chernobyl! Three Mile Island! Fukushima! Look how bad nuclear power is! Give us money, we might kill everyone!", it's more like "We're small and safe!"

Teller attempted this strategy, and nobody liked the idea, even in the context of it being an un-preventable response in the context of MAD as a deterrent.

> Nvidia got big because they're an actual monopoly, the sort of demand-driven business that OpenAI and Anthropic envy.

1. Both OpenAI and Anthropic are demand driven. Absurdly fast growth in both cases.

2. Nvidia without AI driving their demand would be less than half their current size.

3. My point was responding to you saying he was doing a bad job of boosting his company.

> Since America's labs can't beat the openness of Chinese LLMs, they want the authority to be a domestic monopoly for "safety" purposes and demand a slowdown to staunch their bleeding in the global market. None of this is voodoo economics,

Founding documents for OpenAI and Anthropic both talked about safety, as do many of the other startups that followed in the wake of ChatGPT.

I wouldn't put it past any of them (not just OpenAI and Anthropic, but also NVIDIA, Grok, Meta, Google, etc.) to want to manipulate politicians of course: half of them in any legal regard including but not limited to regulatory capture, the other half I wouldn't trust to not do crimes for that political manipulation (Musk offering to pay voters, WTF).

However, people at every level from top to bottom? People, from young researchers to actual Nobel prize winners, resigning to say the safety isn't good enough? Third parties getting hacked by systems which at the time it wasn't obvious were competent enough to discover zero-days but which since the attack we now use for that on purpose? Different third parties investigating the hack having a long record of looking for risks and reporting about them?

Compare with X: people find ads shown next to extremist content, report it, advertisers upset and withdraw, Musk SLAPPs the researcher group out of existence and go after the advertisers for (supposedly) unlawful collusion. Releases an image tool that is immediately misused for CSAM, Musk laughs off reports by using it to put himself in a bikini, then makes it a paid-only tool; laws get passed, he fights those laws and claims it's a free-speech case.

Compare with fossil fuels: their own research decades ago showed it was bad for the world, they suppressed the evidence. Similar for cigarettes.

While I still don't know why you think ROI is argument-shaped, my understanding is that if fossil fuel industries had to pay for their externalities, they would all be loss-making.

https://www.cbsnews.com/news/jensen-huang-nvidia-rejects-ai-...