OpenAI chief scientist, Jakub Pachocki: We believe that deep learning may be under a decade from superintelligence — smarter than us across many critical axes The biggest impact: faster science and new technology Likely the most significant long-term effect is speeding the pace of progress

Under a decade? What happened to next year.

The greatest achievement of Altman it's not his LLM derivative software, but the way he scurried from accountability and make the public give him billions for chasing the horizon in a perpetual placebo cycle of copyright infringement. Just a con man, a Mahatma driving Bugatti.

It will be here in a shorter time span if we step away from LLM. LLM is not capable of super intelligence - making this conversation obsolete.

When intelligence itself becomes a renewable resource, progress will compound beyond imagination. The real question isn’t if superintelligence arrives—but whether humanity can evolve fast enough to guide it.

This guy has a big brain you can see it

what a time to be alive 🤗

He is @elonmusk

This is why compute is the new oil. The nations and companies controlling AI infrastructure will control the rate of scientific progress itself - a new form of geopolitical power.

You can’t even have intelligence from deep learning, let alone super intelligence

AI detector algorithm: detectBsAiPromise() { if(yearsOftimeline >=5 && yearsOftimeline <=10) { return "Bs Promise";

We just need models that do 99.9% on SWE benchmarks.

Oh, so basically that approaches is if we give it enough computer it will wake up with a why 1 day that’s cute 🤣

If that timeline holds, we’re closer to rewriting human progress than anyone expected.

this timeline is insane, accelerating science sounds dreamy but what worries me is the governance side, like how do we make sure LLMs don't just break everything, you think open research helps or hurts there?

Superintelligence isn’t a question of if anymore, it’s a countdown to when science itself starts iterating faster than we can comprehend.

This changes everything. We're mapping where AI agents hit their limits and what it means for automation workflows. Understanding today's constraints becomes tomorrow's moat. We explore this daily in our feed—come see how teams prepare workflows for superintelligence scaling.

That's a fascinating perspective on the future of AI and its potential impact on science!

Yeah !you’re right. It’s easy to smarter than you😏

Hold on — did I miss something? I thought today’s AI was already “smarter than us” across countless axes — total pseudo-recall of the world’s info, PhDs in everything, right? The issue isn’t intelligence on an axis. It’s competence in the real world. That’s what we’re solving.