ILIA

Building Enthusiast

[timo @timo_rf]

they named it that because after 4 prompts all your tokens Argon

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

I LOVE THE SMELL OF FREEDOM IN THE MORNING

[Department of State @StateDept]

SECRETARY RUBIO on https://t.co/pt9rQXC44r PASSPORTS: No printing forms, no drugstore hostage photo, no appointment, no line, no middleman that you're paying to do something that should be available to you.

[Milind S @milindlabs]

wow I cannot believe this

https://t.co/4lJaF6C0c1

[Theo - t3.gg @theo]

This is where we were at with AI code early last year btw

[Larry Dean @larry_mcwethy]

Why don't we eat Bull meat?

[ILIA @iliaghp]

@sergioavilax @ClaudeDevs That's what I'm thinking

[ILIA @iliaghp]

Gemini told me it can center a div😔

[Theo - t3.gg @theo]

It's hard to justify using models that are dumber than Fable and Astra. For real-world code work, the benefits massively outweigh the cost.

The benefit isn't better code, it's something else that's more subtle.

[Theo - t3.gg @theo]

It makes me a bit sad that we see this as “Apple innovation” and not the people behind it.

Tell us about the designers that came up with it! The humans deserve more credit.

[Theo - t3.gg @theo]

Despite the reassurance of @_mohansolo and the Deepmind team, it does appear as though many Antigravity users are still getting their Google accounts banned. They did a big wave today.

[ℏεsam @Hesamation]

WAIT… Astra crushes Fable 5.1 on every benchmark but makes 0 progress on the AA Intelligence score?

> same 61 as Sol and Grok 4.6
> 51 on Agentic Index vs 59 for Grok
> and then it saturates ARC-AGI-3?

this is the biggest discrepancy I’ve seen a model’s evals.

[Dev Shah @0xDevShah]

Hey Tom, appreciate the support. But if we're going to invoke “trust,” perhaps you can also explain why did you:

> Ban Windsurf when it was about to be acquired by OpenAI
> Block xAI employees on competitive grounds
> Cut off OpenCode because it was a worthy CC competitor
> Go after ClawdBot, now OpenClaw
> Kill the “Login with Claude” for all third-party harnesses and agents

Can you also care to explain why openness at Anthropic is only celebrated until it creates competition?

Because we are arguing trust, kindly share why Claude’s sub is the most restrictive sub in the AI ecosystem. Users pay for Pro/Max, yet the only allowed tools are subpar Claude Code and sub-subpar Cowork?

[Chubby♨️ @kimmonismus]

Anthropic: we are reducing your weekly rates by 17%.

OpenAI: We are resetting usage for all paid Codex and ChatGPT Work users and fixed several bugs that wasted usage, so the same allowance should now last 10–50% longer.

At this point its crystal clear who does better community work.

[ThePrimeagen @ThePrimeagen]

?

[Chubby♨️ @kimmonismus]

When Windsurf was set to be acquired by a competitor, you ruthlessly cut off access. Now Cursor is being acquired by SpaceX, which also owns xAI, and suddenly you are emphasizing loyalty and even expanding support. Thats just dishonest, Anthropic.

[AndrewAzlan @AndrewAzlan]

@thsottiaux The irony of a company founded as a non-profit to secure funding, which turned private to secure profit, complaining about the guy who provided the funding while it was a non-profit, on his platform, where they've paid for a blue checkmark.

You guys are hilarious.

[tobi lutke @tobi]

I’m thinking about banning Claude code at Shopify until they change their mind and read AGENTS.md and .agents/skills etc.

Insisting on only reading CLAUDE.md sometimes leads to split brain problems when different team members use different tools. Just unnecessary.

[ILIA @iliaghp]

@NonStateFront @CanadaFP For protecting coastlines?🤣 as an Iranian, IRGC has been terrorizing the world, they attached over 10 neighbouring countries out of spite. They’ve been funding proxy groups and terrorist attacks accross the globe. Hezbollah, Hamas, houthis, hashad al shabi, fatemioun…

[Theo - t3.gg @theo]

This comment just sent me down the wildest rabbit hole

[ILIA @iliaghp]

Could it be Google?

[Theo - t3.gg @theo]

“We have capacity for 100T tokens per day”

Okay who the fuck made this model and where did they get this much compute?