Doomers have lost the AI fight

[FergusArgyll]

https://www.axios.com/2024/05/16/ai-openai-sam-altman-illya-sutskever

[richrichie]

It is impossible to regulate any tech that has military value.

[eesmith → richrichie]

Nuclear weapons are pretty well regulated. I'm only aware of nations - the ones who make the regulations - which have them.

[richrichie → eesmith]

Israel, North Korea, India, Pakistan and Iran did not bother about NPT. KSA most likely has nukes as well. Japan can assemble one fairly quickly.

We also have international treaties for chemical and biological weapons. But big countries are known to engage in things like pathogen gain of function research.

[eesmith → richrichie]

Treaties are not regulations.

You might just as well drop the "military" from your statement.

We've got world-wide regulations on CFCs but people break then, because producing CFCs has value.

[chrisjj]

> the doomer scenarios have looked increasingly far-fetched as the public has grown more accustomed to the limits and frustrations of today's available AI tools.

Really? I'd have thought the opposite. The more accustomed the public grows to any aspect of today's "AI" tools, the more likely becomes the threat of tomorrow.