Is This Actually Intelligence? (Claude, ChatGPT, DeepSeek)
Impossibile aggiungere al carrello
Rimozione dalla Lista desideri non riuscita.
Non è stato possibile aggiungere il titolo alla Libreria
Non è stato possibile seguire il Podcast
Esecuzione del comando Non seguire più non riuscita
Is This Actually Intelligence? (Claude, ChatGPT, DeepSeek)
-
Letto da:
-
Di:
Three AI models, fifteen questions, one conversation that never actually happened.
I put the same fifteen questions to Claude (Anthropic), ChatGPT (OpenAI) and DeepSeek — same order, each in a separate session, none of them aware the other two were being asked the same things. Not a competition. A pressure test.
The idea behind the series, Coordination Capital, is that AI should be measured by the work it finishes, not the tokens it burns. That only holds if there's something like judgment on the other end. So I asked, fifteen ways, whether there is — and watched where the three of them split: on whether there's anything it's like to be them, on whether their perspective is real or a convincing simulation, and on what these systems are actually worth. One of them didn't defend the framework or dismantle it. It replaced it.
I went in to prove a point and came out less certain than when I started.
The Post-Project World — how we govern and pay for work in a world where AI can complete it.