AI Research
AI's Thinking Problem: Sapient's Latent Leap
We've been giving LLMs scratchpads to 'think,' but what if the real reasoning happens *inside* their black boxes? Sapient's new model suggests a radical alternative.