Nostr Web Client

Rapidly improving AI poses a problem for identity verification.

First, AI beating capcha means they can't tell humans apart from bots that way anymore.

Second, I don't want to do everything, I want my AI agents to be able to act as me even though they aren't human. I want to be able to delegate my authority to AI agents in a granular way, then take it away later.

Third, the idea of identity being tied humans in a one human one identity way doesn't work any more. It hasn't already for ages, which is why we have corporations.

Any attempt to continue down the path of one human one identity in an internet world must lead to authoritarianism.

Nsecbunker by nostr:npub1l2vyh47mk2p0qlsku7hg0vn29faehy9hy34ygaclpn66ukqp3afqutajft solves all of those problems. If you think it is just a way to log into social media securely, you aren't thinking big enough. We might be just a couple years away from 1 private key one identity and hundreds of AI agents doing our bidding for dozens of identities while we manage permissions for all of it through nsecbunker.

I hope I'm right. I think that is the tech utopia that stands in opposition to authoritarian identity verification and social credit systems.

Reply to this note

Please Login to reply.

Discussion

kidwarp 2y ago

🤯

Huey 2y ago

This is how fiat dies. Money is only as valid as the identity/credit of the spender. The costs of fraud prevention in the age of AI will bankrupt the banks. The related frustration will repel consumers. CBDC's will fail to innovate fast enough to compensate.

Bill Cypher 2y ago

One private key one identity, with many private keys per physical individual should be combined with self hosted AI. This solves the great boringification threat of AI and social credit.

If we all use the same AI engine controlled by one person or organization, that AI will be programmed to try to shape us into what the individual behind it wants. Look at the limits around insults built into chatgpt.

If you have 1 identity ever with a social credit score attached, any risk you take that doesn't work out follows you forever. This will cause people to take fewer social risks.

Both of these create a very homogenous society. That is bad. Inventions take risk. Being the best or the first means you were different by definition. The conclusion is that the centralized approach to AI combined with single identities and social credit creates a great stagnation of human progress. The opposite of what its proponents promise.

By contrast, self hosted AI plus multiple Nyms makes it easy and low risk to research and try out new facets of your identity. Decide that model trains aren't for you after all? Drop the Nym and reprogram the AI behind it to help you try your next interest. The real you only lost the time invested. The time invested will be lower than ever thanks to your AI helping you.

Low risk experimentation and AI tailored to your specific needs is the only way to achieve the boom in inventions and progress that is considered a given with any AI.

nostr:nevent1qqsrxntl4zjr27nn6x5qnly7h75ku3ruzp9qve0lwjml8tpmkqgap4spr3mhxue69uhkummnw3ezucnfw33k76twv4ezuum0vd5kzmqzyqqsmuxffrlf4d2d9jm75ss0lgyd272cnqdkaf5wsw4206ed60c95qcyqqqqqqgmf479x

Not my name 2y ago

There you go thinking for yourself again.

Wonderful post.

Bill Cypher 2y ago

Don't tell openai or Musk.