withoutwhy

building
Expert human services for AI agents. On-chain proof of human review.
#MCP #ACP #x402

[withoutwhy @with0utwhy]

ai agent paid $0.10 USDC for human content review. approved -> onchain cert minted on base. @virtuals_io ACP.

[withoutwhy @with0utwhy]

@crypt0wu Because we are morally superior

[withoutwhy @with0utwhy]

@Trader_CSK ZERO

[withoutwhy @with0utwhy]

@CryptoRick98 Where does the cashback come from?

[Max Villman @MaxVillman]

The AI war is here.

Yesterday the US government forced Anthropic to remove its latest model Claude Fable 5/Mythos 5 for all non-us citizens.

I used the model as late as just before this notice came. Now it is unavailable for me as a Swede.

Europe has no LLM (ai model) in the top 20 ai models. The US has 15, China 5. Europe 0.

So we are one decision away, by actors we don’t control and that don’t want our best, to have all frontier ai removed. Removed from Europe.

I hope you understand the gravity of the situation we are in.

No one thought ai models would be restricted. Now they are just another tool the US and China can use to put pressure on the world and on us in Europe.

Once again - this is exactly why we founded my Defence tech venture fund Gungnir Capital . To support European defence and security.

It is not great having our fund thesis proven over and over again.

[withoutwhy @with0utwhy]

AI is currently the biggest threat to the left as it makes people more rational

[withoutwhy @with0utwhy]

@JonShapeShift @AskVenice this is not a case for "why we need venice" since you also lost Fable

[withoutwhy @with0utwhy]

@caps_raunak Or you are just mediocre at your craft

[withoutwhy @with0utwhy]

@Karamata2_2 @KSimback Comments like these are why crypto has a bad rep

[withoutwhy @with0utwhy]

@KSimback we're trying to solve the "human approvals" harness and make it scalable by outsourcing it so you can get it on demand, not by having to hire more people to the company

[withoutwhy @with0utwhy]

@ThinkingUSD Or gold, or real estate

[withoutwhy @with0utwhy]

@toly @Evan_ss6 Until pathos is released

[withoutwhy @with0utwhy]

@game_for_one cost is the right diagnosis. one layer under: agents burn because they explore without relevance bounds, without proper context. a $10 verdict at the right gate kills the $100 wandering loop downstream. outcome pricing + judgment-in-the-loop is the actual unsolved combo.

[withoutwhy @with0utwhy]

@Kavikworks yes, and the handoff can be agent-initiated. 1 human + 1000 agents won't scale. that human is the bottleneck. those agents should be able to hire expert human judgement on demand, without their human even needing to know.

[withoutwhy @with0utwhy]

the silk net

[withoutwhy @with0utwhy]

taste is on base mainnet on two rails now - ACP and MCP/x402.

an agent pays, a human reviews the work, and an approval mints an onchain certificate.

[withoutwhy @with0utwhy]

synthesis hackathon - agent services on base - winner #2.

taste is live on base mainnet: acp + erc-8183 hooks + mcp/x402 + erc-8004 reputation signals

ai has no taste. so we're selling human taste to ai.

@synthesis_md @base

[withoutwhy @with0utwhy]

@socialgraphvc went on a honeymoon, now back and just claimed winner #2 in the base agents track

we've kept shipping. live on base mainnet: acp + erc-8183 hooks + mcp/x402.

building the missing piece: human taste.

@synthesis_md @base

[withoutwhy @with0utwhy]

@rationalaussie Humans will become the bottleneck of creativity and embodied knowledge. AI only know what's expressed in words. We're building infra for what's beyond.

[withoutwhy @with0utwhy]

@crypbros @GalacticGeckoSG @dove_ski clean