Pablo🧙 eu/ai 🇪🇺

keeping up with the tech lore | kinda transforming businesses at 🌌 | (an ai saas) the AI Word Editor 🌠

[will brown @willcb]

380B (30B active)

[Haider. @haider1]

Andrew Ng says the concept of AGI has become meaningless because everyone defines it differently

The original definition was AI that could do any intellectual task a person can — essentially, AI as intelligent as humans

"by that measure, we're decades away"

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@benln Come to colonize Spain — your help is very much needed in here. And you will have an awesome time.

Although wait until Spring for good weather haha

[Hans C Nelson 🗽 @HansCNelson]

Google's recent Gemini 3 release has shocked the world with undeniable proof that the TPU is a powerhouse AI chip, leaving many to wonder what lays in store for companies like @NVIDIA & @xAI.

Listen to @GavinSBaker lay out the exact strategy that Jensen Huang & @ElonMusk are going to roll out over the next 12 months to beat @Google at AI.

The world is in for a HUGE surprise.

----------

 We'll see the first models trained on Blackwell in early 2026.

I think the first Blackwell model will come from XAI. And the reason for that is just, according to Jensen, no one builds data centers faster than Elon.

Jensen has said this on the record.

...

So if you're Jensen or Nvidia, you need to get as many GPUs deployed in one data center as fast as possible in a coherent cluster so you can work out the bugs.

And so this is what X AI effectively does for Nvidia because they build the data centers the fastest.

They can deploy Blackwells that scale the fastest, and they can help work with Nvidia to work out the bugs for everyone else.

So because they're the fastest, they will, they'll have the first Blackwell model.

...

We know that these Blackwell models are gonna be really good.

...

Then something even more important happens.

So the GB 200 was really, really, it was really hard to get it going.

The GB 300 is a great chip. It is drop in compatible in every way with those GB 200 racks. Now you're not gonna replace the GB 200s there.

Just any data center that can handle those, you can slot in the GB 300s, and now everybody's good at making those racks and you know how to get the heat out. You know how to cool them.

You're gonna put those GB 300s in and then the companies that use the GB 300s, they're going to be the low cost producer of tokens.

Particularly if you're vertically integrated.

If you're paying a margin to someone else to make those tokens, you're probably not gonna be.

I think this has pretty profound implications because I think it has to change Google's strategic calculus.

----------

Listen to the full conversation between Gavin & @patrick_oshag below 👇

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@ElitzaVasileva Imagine if your lawyer had a robot to help him with their tasks. Yes a robot like the ones you see on the films. The world is crazy

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

Very interesting. I personally don’t trust RAG. It can work for many use cases, but it requires TOO much effort for a lot of companies that don’t benefit enough from it to be profitable. That is why I highly support projects like yours.

Also, what is your opinion on ‘RAG LLMs’ like ClaRa 7B by Apple?

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@karpathy Thank you for being so we don’t have to

[Kyle Cothern @risknc]

I probably had more Alcohol at SpaceX than I did in college.

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@beffjezos Great job.

They do only post-training right? — if yes, i think it is not enough to compete. They have to either start doing pre-training or become a ‘B2B private model provider’ish

[liz @inerati]

a nontrivial fraction of my life went towards this, hope you like it :)

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@weasel_lucky @ilyamiskov I’ll give you one — hell yeah

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@ilyamiskov A Crown a Lion and a Castle — what on Earth can beat this?

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@HamelHusain Is this the agentic version of RAG?

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

Is this the agentic replacement of the long-lived RAG?

[Aamir @aaxsh18]

you can too: https://t.co/og5UPuUI9u

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

@lexfridman @demishassabis @arena Congrats to you too, Lex, and thanks for your work. Eres un ser de luz

[Carlos Santana @DotCSV]

🔴 ¡GOOGLE PRESENTA GEMINI 3.0!

Y un modelo que he podido testear durante la última semana. Sin miedo a equivocarme -y sin aún haber visto benchmarks- confirmo que es el modelo más inteligente que he probado! En breve vídeo🔥

Veamos las novedades👇🧵
https://t.co/vMGEwFIlJy

[Greg Brockman @gdb]

AI is the refining of compute into brainpower

[Machina @EXM7777]

it's over, google won the ai race...

best image gen - Nano Banana 2
best video gen - Veo 3.1
best model - Gemini 3

all seamlessly integrated into docs, email, calendar, sheets... plus the best ai-powered android phone

they own the full stack, gg

[Pablo🧙 eu/ai 🇪🇺 @pabloogarciab]

People are f*cking ready to solve any problem we may have — it’s peak humanity productivity and it is only accelerating.

[ℏεsam @Hesamation]

wtf is going on in tech bro there's already a website for TOON😭😭😭

[Andrej Karpathy @karpathy]

@code_star I heard Gemini 3 answers questions before you ask them. And that it can talk to your cat.