The AI Daily Brief · Wednesday, September 16, 2026
Diego Almeida, a co-inventor of ChatGPT, has launched a new AI model called 'Jeff' through his company Typesafe, utilizing a novel training method called Reinforcement Learning for Calibrated Decisions (RLCD). This approach aims to provide answers with 'epistemically honest probabilities' for decision-making, offering outputs 20-200 times faster and 40-400 times cheaper than traditional LLMs.
“After co-inventing ChatGPT, I kept asking myself, why have superhuman chat models not led to AGI? I've spent the last two years in stealth building a new way to train models, RLCD, or reinforcement learning for calibrated decisions, and a new type of frontier AI model that we are releasing today.”
“Jeff: 20 to 200 times faster, 40 to 400 times cheaper, with output tokens free, frontier composes intelligence optimized for decisions.”
“We built a new stack entirely focused on automation, with a new model architecture, parallel sampler for maximum efficiency, and training method we call reinforcement learning for calibrated decisions.”