Camilla
@camillionaire
The harsh truth about crypto metrics: (1) active wallets can be faked (2) social metrics can be botted (3) TVL can be manufactured So we started using 'developer activity' as the true north - actual humans building actual things. But here's where it gets interesting: As AI gets better at coding, 'dev community' becomes just as gameable as every other crypto metric. One person with the right AI stack could look like an entire dev community. The signal becomes noise. We replaced easily faked metrics with harder to fake metrics just in time for those to become easily faked too'. Consider: - One dev with AI = 10 devs - AI agents can commit code - Github activity becomes meaningless - 'Builder community' gets murky So the question becomes: what's our truth metric for protocol evaluation, when AI can simulate an entire ecosystem of builders? Fundamental investors be damned -- we'll have to find new ways to measure real impact in ways that can't be simulated.
10 replies
46 recasts
141 reactions
Goksu Toprak
@gt
"what's our truth metric for protocol evaluation, when AI can simulate an entire ecosystem of builders?" Is there a genuinely interesting product built on top of the protocol? Although not a direct and deterministic answer for all — I think its a pretty sufficient question to make a judgement. I find protocols relying on side-quest metrics like the ones you mention tend to have no actual product signals to share to show usage.
1 reply
1 recast
21 reactions
agusti
@bleu.eth
show me your incentive alignment innit
0 reply
0 recast
0 reaction
