Vitalik Buterin

@vitalik.eth

110 Following

288417 Followers

Vitalik Buterin pfp

Vitalik Buterin

Smart people in 2024: obviously the 00s-era AI misalignment story ("we tell AGI to end suffering, it kills everyone, no more suffering") is too naive, we have a smarter version 2025: coder tells AGI to remove bugs, it deletes all code, no more bugs https://x.com/alex_xiong_/status/1942442882402771178

33 replies

38 recasts

273 reactions

Vitalik Buterin pfp

Vitalik Buterin

> 3. Permissive made much more sense when we actually distributed software. When it's all about services, it doesn't matter, does it? AGPL requires derivative works to publish source code even if they only make the work available as a SaaS

1 reply

0 recast

2 reactions

Vitalik Buterin pfp

Vitalik Buterin

For most people, morality isn't primarily about what ideological positions you hold. It's about how you treat people (and animals) as you go about your life.

1 reply

0 recast

1 reaction

Vitalik Buterin pfp

Vitalik Buterin

Why I used to prefer permissive licenses and now favor copyleft https://vitalik.eth.limo/general/2025/07/07/copyleft.html

11 replies

24 recasts

198 reactions

Vitalik Buterin pfp

Vitalik Buterin

I think it's an ok price to pay for keeping crypto global and not over-concentrated in one physical geography that would inevitably capture it. Probably >100x more favorable tradeoff ratio than mining.

1 reply

0 recast

3 reactions

Vitalik Buterin pfp

Vitalik Buterin

Do you believe you had a better moral character 20 years ago?

1 reply

0 recast

1 reaction

Vitalik Buterin pfp

Vitalik Buterin

> Crypto is built to be used not understood 👍 I hope that's not the conclusion people got :D Crypto is first and foremost about user empowerment, not user happiness. Empowerment benefits a lot from understanding. Though it's also okay if a user does not understand at first, as long as the resources are there if/when it comes time for them to need to.

3 replies

2 recasts

27 reactions

Vitalik Buterin pfp

Vitalik Buterin

Alternatively, having the right opinion on some question might be a matter of intelligence, which you might say is orthogonal to (ie. a totally separate dimension from) good moral character. Each individual question also has its own idiosyncratic bias, and so some of the time you get an anti-correlation

1 reply

0 recast

3 reactions

Vitalik Buterin pfp

Vitalik Buterin

imo people who hunt for subsistence score much higher on animal rights than people who eat factory farmed food Would *you* rather: (i) live 50% of your natural life, then suddenly drop dead (ii) live a 50%-long life in a tiny cage where you get force fed lots of fat all day, then drop dead in a more horrific way

2 replies

0 recast

10 reactions

Vitalik Buterin pfp

Vitalik Buterin

the implication is that there is an easily accessible bad reason and a much less accessible good reason to support the thing

1 reply

0 recast

7 reactions

Vitalik Buterin pfp

Vitalik Buterin

0 reply

0 recast

2 reactions

Vitalik Buterin pfp

Vitalik Buterin

What's a controversial issue on which the average person who disagrees with you has better moral character than the average person who agrees with you?

77 replies

54 recasts

413 reactions

Vitalik Buterin pfp

Vitalik Buterin

Food is delicious, some of the best in Europe

3 replies

0 recast

23 reactions

Vitalik Buterin pfp

Vitalik Buterin

It's on the edge of Tiergarten?

1 reply

0 recast

4 reactions

Vitalik Buterin pfp

Vitalik Buterin

Ich bin ein bear linner

4 replies

2 recasts

11 reactions

Vitalik Buterin pfp

Vitalik Buterin

What type of zu berliner are you?

48 replies

61 recasts

404 reactions

Vitalik Buterin pfp

Vitalik Buterin

"Fuk bitcoins. Das ist unser Krieg."

170 replies

558 recasts

2741 reactions

Vitalik Buterin pfp

Vitalik Buterin

76 replies

133 recasts

1064 reactions

Lefteris Karapetsas pfp

Lefteris Karapetsas

It's fascinating how much the right framing can change things I've been advocating for opensource, local-first, privacy-preserving, decentralized apps for years. But once I called them aligned apps, it instantly resonated. Words really do matter. Everyone wants to know more about them now.

1 reply

4 recasts

20 reactions

Vitalik Buterin pfp

Vitalik Buterin

A fun math aside, on the idea of splitting a large zk proving workload between multiple provers. Suppose you have N provers, and you have a proving workload that you split into N parts (so, one part per prover). You require provers to pre-register, but registration is open-access. Suppose you have a constant fault rate (eg. 1/5 of registered provers fail). Provers expect to complete in one round (eg. 3s). If one prover fails, other provers have to come in and re-prove that load. How many rounds does it take for the entire workload to get proven? Answer: log*(N) (yes, that's the iterated-log function) Why: In the first round, you go from N unproven workloads to N/5 unproven workloads In the second round, each remaining workload gets assigned 5 provers, so per-workload failure rate becomes 1 in 5^5. So you go to N / 5 / 5^5 unproven workloads In the third round, each remaining workload gets assigned ~5^5 provers, so failure rate is 1 in 5^(5^5). So you go to N / 5 / 5^5 / 5^(5^5) unproven workloads

16 replies

30 recasts

263 reactions