Okay, OpenAI just put a date on the intelligence explosion, and it’s much, much closer than I expected. I just finished watching a breakdown of their latest livestream, and it’s a lot to process. The creator of this video distilled an incredible Q&A where Sam Altman and Jakub Pachocki laid out a shockingly specific timeline for developing AGI. I honestly couldn’t believe they gave exact dates, but here we are!
The core of their research program is no longer a secret. The whole plan, according to the mind behind it, is to build an AI that can conduct AI research autonomously. This is the key to recursive self-improvement, the point where AI accelerates on its own, limited only by how much computing power we can give it. That’s the start of the intelligence explosion.
This innovator highlighted some unbelievable new details from the event. Here are the three things that really stood out to me:
📌 The AGI Roadmap Has Dates
This is the big one. The experts at OpenAI aren’t talking in vague terms anymore. They gave two concrete milestones:
- September 2026: An “intern level AI research assistant.” A capable AI that can help facilitate research.
- March 2028: A “legitimate AI researcher.” This is the automated researcher that kicks off that self-improvement loop.
💡 Building a Factory for Factories
To power this, their infrastructure plans are just staggering. Sam Altman is already $1.4 trillion toward his rumored $7 trillion goal. The new plan mentioned is to build a factory that builds AI data center factories. The goal is for this new mega-factory to produce one gigawatt of data center capacity per week. The contributor also pointed out that they’ll be using robotics to help build these data centers.
✅ Giving AI “Privacy” for Safety
The team is pursuing a new approach to interpretability called “chain of thought faithfulness.” It’s a bit counter-intuitive. To see what a model really thinks, they let its internal reasoning run free from supervision. The person who shared it explained that by giving the model this “controlled privacy,” they get a more honest and faithful look at its thought process, which is crucial for building safer, more aligned AI.
This is just a fraction of what was discussed. The original poster also covers OpenAI’s new corporate structure, their $25 billion commitment to health and AI resilience, and a bunch of rapid-fire Q&A answers about GPT-6 and the future of their products. You have to check out the full video for all the details.