Vitalik Buterin on Warpcast

Vitalik Buterin pfp

Vitalik Buterin

Hypothetically, would you rather have: 1. AI is not delayed, but the strongest AI, 10 years ahead of everyone else, is in the hands of a single very powerful multinational inter-governmental agency 2. All AI delayed by 10 years

36 replies

21 recasts

147 reactions

ʞɔɐſ pfp

Vitalik I’m confused. Why cast this 3 times? Bug? Anyways, definitely 10 years delayed; for reasons that seem obvious

1 reply

0 recast

1 reaction

Colin Johnson 💭 pfp

Colin Johnson 💭

If you ever want to deliver surveys like this through @survey , let me or @ba know. We can now deliver survey questions on demand along with visualized, aggregated results when complete 👍

2 replies

0 recast

3 reactions

Mac Budkowski ᵏ pfp

Mac Budkowski ᵏ

2. though it breaks my heart thinking about all the people who could get educated, healed and inspired by AI assistants

0 reply

1 recast

2 reactions

Ayan pfp

I don’t think it is intentional, but creating hypothetical scenarios with false binaries like this can make it challenging to engage with AI ethical tech discussions in a meaningful and non-polarizing way This feels very American politics…👀

0 reply

0 recast

1 reaction

Branksy Pop pfp

Choose 2 and you'll end up with 1. Those who will regulate it will be the ones who profit before the regulated.

0 reply

0 recast

1 reaction

yesyes pfp

2 I've answered 2 in all the cases though I was leaning a bit towards 1 in this one. I don't think that a lot of people realise how powerful an AI that is 10 years ahead would be. At that point it would be practically impossible to catch up to it. OpenAI is like a year or 2 ahead of everyone and the gap hasn't closed

1 reply

0 recast

1 reaction

YB pfp

I only thought about this for 2 mins but my gut says #1. Public key cryptography was first discovered by GCHQ but Hellman & Diffie came to the same conclusion 5 years later anyways. My gut says that if a government has some AI that's so strong, the public will come to it sooner than later as well

1 reply

0 recast

1 reaction

shazow pfp

2 most likely, but could be convinced otherwise depending on the governance of 1 and what "in the hands of" means.

0 reply

0 recast

0 reaction

jovian.eth pfp

1. I have this weird notion that we should rip the bandage off quickly. I'm more concerned about a narrow form of AI in the hands of an idiot than a general recursive AI that reaches escape velocity. Reason being is that I believe a God-like AI would be comparison make *us* look robotic and uncaring.

0 reply

0 recast

0 reaction

🤷🚂👲🧑‍🤝‍🧑 pfp

🤷🚂👲🧑‍🤝‍🧑

2, absolutely. I do not see a very powerful multinational inter-governmental agency leading to materially good consequences, more likely leading to extrajudicial arbitrage, expropriation, and rendition.

0 reply

0 recast

0 reaction

GIGAMΞSH pfp

0 reply

0 recast

0 reaction

Salvino Armati ↑ pfp

Salvino Armati ↑

2. feels like the risks of giving a small group of people (who already have a monopoly on violence) infinitely available intelligence as a cornered resource is far greater than superlinear GDP growth being 1 decade behind

0 reply

0 recast

0 reaction

TheModestThief🎩 pfp

TheModestThief🎩

Probably 1, because option 2 doesn’t preclude a similar fate of the most powerful AI being in the hands of the powerful agency?

0 reply

0 recast

0 reaction

Yix pfp

0 reply

0 recast

2 reactions

187 pfp

@arthurfleck.eth

Can someone explain why 2 is a good idea? We have seen what regulation does (cc bill gurley all in episode). Feel like it helps the select few Sam Altman / Elon etc a lot more than everyone else. Do they not just keep working in the background and then become a massive leap ahead ?

0 reply

0 recast

1 reaction

shoni.eth pfp

0 reply

0 recast

1 reaction

web3magnetic.eth 🎩 pfp

web3magnetic.eth 🎩

Imo, 2 is better than 1

0 reply

0 recast

1 reaction

Serhii pfp

Hm... It seems to me that the first option is better, since one high-quality product stimulates competitors to develop, not to sit still. Even if it is so powerful, analogues will still be built. Healthy competition should generate strong competitors that benefit the end user of the product.

0 reply

0 recast

0 reaction

streetphotography.eth pfp

streetphotography.eth

1 reply

0 recast

0 reaction