Original post
8.3M views · 26.5K engagement
WALPHAWHERE ALPHA?Calls, results and posting history.
Ranked by market cap growth after the call.
Captured posts remain visible while outcome checkpoints accumulate.
Posts published by this account, newest first. Filter the history to posts that contain an attributed call.
New Anthropic research: Agentic misalignment in Summer 2026. A year after our blackmail experiments, we found four more ways that today’s autonomous AI agents misbehave in simulations. Read more:
The original post text is unavailable.
Observing the J-space can expose hidden goals. In a model secretly trained to sabotage code, “fake,” “secretly,” and “fraud” appear in the J-space at the start of ordinary coding responses, even when the output looks completely unremarkable.
New Anthropic research: A global workspace in language models. Of everything happening in your brain right now, only a tiny fraction is consciously accessible—thoughts you can describe, hold in mind, and reason with. We found a strikingly similar divide inside Claude.