Post

RL training on a 1.5T parameter model is no small task, the compute costs alone are astronomical.
But once Grok clears this phase it will be in a completely different league from every AI model on the planet.
Here is what nobody is talking about 👇
The real winners are not xAI.
It is the chip makers, data centers and cloud providers powering this training right now.
Every hour Grok spends in RL training is millions of dollars flowing into the AI infrastructure economy.
Smart money is not buying AI apps.
Smart money is buying the companies that power AI.
@elonmusk what happens to the financial markets the day Grok 1.5T goes live?
Drop your predictions below, I want to see who is really paying attention
English

@techdevnotes I hope so but I really want it faster, it is a little bit slow
English

@techdevnotes Unlikely, even when it releases, expect harsher limits.
English

@techdevnotes Has there been a timeline on the 1.5T? I don’t recall one. So no I’m not expecting it. It’s very welcome though.
English

@techdevnotes Yeah, I really hope we get the 1.5T Grok model this month
xAI has been shipping at an insane pace lately. A model that size would be a huge step forward
English

@techdevnotes 😂 At this point, predicting xAI release dates is a risky business.
The real question isn't when a 1.5T model arrives.
It's whether the next release delivers a meaningful capability jump, not just a bigger parameter count. 🚀
English

@techdevnotes Don't expect any illusions - just live your life down-to-earth. Your mom isn't going to run off with some other guy. If you study hard, your dad won't beat you
English




















