artificial intelligence
We borrowed the wrong word. Hallucination implies a sober baseline that occasionally breaks down. But an LLM has no knowing and no not-knowing. It only has one process. The Oracle never knew either. Neither when she was right, nor when she was wrong.
artificial intelligence
And why that’s very bad for agents
artificial intelligence
And why maybe we’ll be safe even if AI becomes more intelligent
artificial intelligence
You Have Never Been Understood. Neither Have I. And That’s Fine (For Now.)
tech
A reflection on enterprise AI adoption, risk ownership, and decision paralysis under accelerating technological change
tech
When statistical models meet subtle human creativity
tech
Why the hardest part of self-driving isn’t the code — it’s the second before we take the wheel.
tech
The unintended consequences of rewriting the balance of power between apps and the few platforms still standing — the operating systems
tech
From fool-proof to fast-fix — and the quiet decay of engineering conscience
tech
From highways to bandwidth to GPUs: the same story, rewritten in silicon
tech
We humans like stories. They help us make sense of life. We want to know why things happen, what caused what, and where it all leads. One…
tech
What's your thoughts?