J.T. Lorentz

Writer and health enthusiast

[J.T. Lorentz @jtlorentz]

@theo @sama Weird, you being a codex fanboy and all

[J.T. Lorentz @jtlorentz]

@jayair So what the f is it about sandboxing that is so hard to contain an LLM?

Is defining a perimeter hard to do? Is enforcing the boundary tricky?

I wanna know about this seemingly uncapable *sandbox*

[J.T. Lorentz @jtlorentz]

@theo What is it about T3 that has given you this advantage?

[J.T. Lorentz @jtlorentz]

@theo This only shows me that I could ship faster if I were more organized.

Thabks Theo

[J.T. Lorentz @jtlorentz]

@cmuratori But what about our precious singularity :(

[J.T. Lorentz @jtlorentz]

@bryan_johnson Are u ragebaiting us lil nigga

[J.T. Lorentz @jtlorentz]

Lmmffaoooo

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

😵‍💫

[J.T. Lorentz @jtlorentz]

Neeeerrrrrrrrrdddddddddd

[ThePrimeagen @ThePrimeagen]

@Wronglebowsk sorry, i honestly have no idea anything about sportsball

[J.T. Lorentz @jtlorentz]

@Perpetualmaniac @theo I was going to ask, is there not a way for gpu usage to be logged as such (or build a tool to track it) ?

[J.T. Lorentz @jtlorentz]

@AshtonForbes @Gt5Skier @JakeHunter88 Would you say mass is a blip or a pimple in the aether and since aether always seeks equilibrium, it tries flattening mass back in?

So gravity is compression inward, not mass pulling things in..?

[J.T. Lorentz @jtlorentz]

Never seen a more perfect reaction holy

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

@Polymarket

[J.T. Lorentz @jtlorentz]

This is what AI psychosis looks like

Senior devs having raging boners for small tools they develop for themselves and completely forgetting about shipping the main thing they're working on.

The paperclips will come if they're shiny enough.

OR, Prime loves small tools.

Hmm..

[ThePrimeagen @ThePrimeagen]

The AIs are playing my game... losers

THE AI EVEN PLAYED 2 DAMAGE ENHANCEMENTs AND ATTACK SPEED... LFG!!!

[J.T. Lorentz @jtlorentz]

Would be great to get an update on the experiment with the same-prompting different LLms + yourself comparison you did.

Did you finish that or..?

I am still waiting ;(

[ThePrimeagen @ThePrimeagen]

ok fella's

i am going to try to loop tomorrow

that is it
that is the tweet

i'll report back by Monday or Tuesday, but i have feeling like its more like 3 weeks from now

[J.T. Lorentz @jtlorentz]

@avrldotdev Local first disk memory

For CRUD app

For performance and UX

[J.T. Lorentz @jtlorentz]

@ThePrimeagen @zeeg Is tab complete the thing you built that one time?

Or just the usual autocomplete in IDEs?

Im nob pls tell me

[J.T. Lorentz @jtlorentz]

Ghotz made a huge bet and its becoming reality.

[the tiny corp @__tinygrad__]

I have on good authority that GLM 5.2 is running at 120 tok/s across two networked Blackwell tinyboxes. $150k and that setup can be yours, either 2x tinybox or 1x tinybox pro. Never pay the cloud again.

[J.T. Lorentz @jtlorentz]

@GadSaad @megynkelly OR invest and then leave.

[J.T. Lorentz @jtlorentz]

@hiarun02 @realpvarma Design outsourcing too

[J.T. Lorentz @jtlorentz]

@avrldotdev @devXritesh Why

Asking bc im noob

[J.T. Lorentz @jtlorentz]

@ThePrimeagen @cheicolate I opened X to see the results.

I wanna see Prime. LET ME SEE.