Jod

Probably right. Definitely questionable.

[Jod @KingJod]

Donate a meal to a needy family.

I gifted a meal by taking a photo at @scooterscoffee with the free @GiftAMeal app.

#ScootersCoffee #GiftAMeal https://t.co/Jh7UD4MkyH

[Ayarsız Astrolog @kediastrolog]

Çin bebek ürünleri konusunda çok ileride 😳

[Jod @KingJod]

@LauraLoomer Anything can happen in America and everyone automatically thinks its staged....so annoying.

[Jod @KingJod]

Literally nothing can happen in the world today that isn't a "staged event"....isn't there something said somewhere when you can't tell whats real and whats fake then here comes the end days?

[Jod @KingJod]

everyone scroll X and find any accounts with just one post, and just a name, I'm trying to see something.

[Jod @KingJod]

You’re treating intelligence like it requires perfect originality and zero blind spots. Humans do not meet that standard either. The question is not whether AI will magically know what it does not know. The question is whether it can build systems that expose its errors, test new ideas, and improve through feedback faster than humans can. That is exactly how human science works, just slower.

Also, calling AI “statistical regurgitation” ignores examples where AI systems have discovered strategies and algorithms humans had not found, like AlphaTensor in matrix multiplication. That does not prove superintelligence is here, but it does disprove the idea that AI can only remix text from training data.

The strongest argument is not “AI definitely becomes god.” The strongest argument is that once AI can automate enough of research, coding, testing, and model improvement, the ceiling may stop being human creativity and start being compute, feedback loops, and verification. That is a very different world.

[Jod @KingJod]

@sama Awww, we love you too @sama

[Jod @KingJod]

There’s so many cry babies on X

[Jod @KingJod]

@Kekius_Sage I think of it as when we have our own kids and the emotions that come with it.

[Jod @KingJod]

I've used @ChatGPTapp since the beginning and @sama you've done a great job with it so far.

[Jod @KingJod]

make me a visual time line of AI in the united states in the next 10 years and the advancements if the pace continued and fastened. Made with @grok

[Jod @KingJod]

make me a visual time line of AI in the united states in the next 10 years and the advancements if the pace continued and fastened made with @ChatGPTapp

[Jod @KingJod]

@nikitabier why they hate on you so much?

[Kim Dotcom @KimDotcom]

Petition to fire Nikita Bier from X

Please like if you agree

[Jod @KingJod]

@nikitabier This is cool

[Jod @KingJod]

@sama @sama 5.5? 🙏🏽

[Sam Altman @sama]

Really excited for this week!

Next up, we've got something to show you at 12 pm PT today.

[Jod @KingJod]

Yes, human polymaths exist. Yes, interdisciplinary scientists exist. But they are still bottlenecked by human memory, human speed, human lifespan, human coordination costs, and human attention.

That is the point.

The claim is not that “mixing domains = superintelligence.”

The claim is that a system that can integrate vast numbers of domains, hold them active at once, test combinations rapidly, and iterate continuously without normal human bottlenecks can cross into something functionally beyond human cognition.

Leonardo was extraordinary.

But Leonardo could not ingest all published physics, biology, code, economics, engineering, history, and lab results in days, then recombine them instantly.

No human polymath can.

So this is not “moving the goalposts.”

It is recognizing that capability can become superhuman through scale, speed, parallel synthesis, and persistence, even before some mystical alien reasoning appears.

And on “superintelligence should perform beyond human capacity in any given task”:
that standard is too absolute to be useful.

A system does not need to exceed every human at every single task to become civilizationally dominant.
It only needs to outperform humans across enough economically, scientifically, strategically, and organizationally important tasks, consistently and at scale.

That would already be a practical superintelligence, even if a few edge cases remain human-favored.

[Jod @KingJod]

@LauraLoomer @RealCandaceO See, God doesn't want you saying that.