ChatGPT Co-creator Unveils Jev: A New AI Architecture for Autonomous Driving
ChatGPT Co-creator Unveils Jev: A New AI Architecture for Autonomous Driving
Diogo Almeida, a co-creator of ChatGPT, has introduced Jev, a novel AI model that processes data in parallel rather than generating sequential tokens. The technology is already showing promise in autonomous driving simulations.
For two years, Diogo Almeida, one of the creators of ChatGPT, secretly developed Jev — an AI model that does not write text or generate tokens sequentially. Instead, it takes any unstructured data and processes it in parallel.
Jev is already being tested in autonomous driving.
According to the author of the experiment, in less than an hour, the new model managed to assemble a full decision-making cycle for a car in simulation: turns, lane changes, crossing intersections, and merging onto highways. For each action, the model returns not text, but typed probabilities of possible decisions along with a confidence level.
This is the key difference between Jev and standard chat models. The input is the system state, and the output is a set of permissible actions and their probabilities.
If this approach scales well, the barrier to entry for creating autonomous driving systems could significantly decrease.
Why it matters
AnalysisMoving from sequential token generation to parallel action probability analysis could revolutionize the development of autonomous systems, making them faster and more predictable.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
Autonomous AI Operator on Telegram 24/7: Architecture, Telethon, Security, and 10 Production Scenarios
A comprehensive engineering guide to building a production-ready AI operator on Telegram using Telethon: session isolation, confirmation systems, ban protection, and 24/7 deployment.
Loop Engineering: Architecture of Autonomous Agentic Loops and the Post-Prompting Era
A deep dive into Loop Engineering: self-sustaining execution loops for Claude Code and autonomous agents, the DOER/CHECKER topology, deterministic verification, the /goal primitive, and reward hacking mitigation.
AI Agents (Autonomous Agents)
An autonomous system based on a large language model that not only responds to messages but independently plans a sequence of actions, utilizes external tools (browser, terminal, databases), and executes complex tasks without constant human oversight.
AI Agents (Autonomous Agents)
Software systems based on LLMs that can autonomously perceive the state of the environment, decompose complex goals, invoke external tools, and iteratively correct their own mistakes.