Back to explore

Testing TypeSafe's Jev on blackjack

A user shares an experiment testing TypeSafe's Jev on blackjack: context and typed questions in, probabilities out, with code doing the arithmetic, reaching 98% over 442 decisions.

Ran an experiment on @typesafeai's Jev, tested on blackjack where every play has a provably correct answer. Not an LLM: context + typed questions in, probabilities out. Code does the arithmetic. Doesn't beat the house edge. But 98% over 442 decisions, the capabilities are clear.

· 0 likesOpen on X