adyingdeath

Everybody, our jobs are coming back.

@awkwardgoogle At the first glance I thought why he naked below in the TV.

He needs to try a new color of pants when on TV

@awsaf49 @thsottiaux Now it's confirmed to be true haha

@thsottiaux openai was right after all lol

I don't know what they are doing over there, but Codex will continue to be available both in the FREE and PLUS ($20) plans. We have the compute and efficient models to support it. For important changes, we will engage with the community well ahead of making them. Transparency and trust are two principles we will not break, even if it means momentarily earning less. A reminder that you vote with your subscription for the values you want to see in this world.

@ghumare64 Since you can force push, you can just do something bigger.

my last day at job.

@IfeanyiEmm34386 Genius student.

Teacher: Why are you late? Student: My dad is in the hospital *7 days later* Teacher: Is your dad still in the hospital? Student: Yes. He's a doctor. Teacher:

@HumansNoContext Bro failed every times but ended up winning.

Task failed successfully

@Akasheth_ claude opus 4.7 improves a lot in consuming tokens compared to 4.6, with a "better" tokenizer costing 1-1.3x for the same prompts.

claude opus 4.7 was fun for about 15 minutes

@DudespostingWs The bro recording this video is a oh my xxx guy.

This blind dude nails the 10-second challenge and earns himself a free haircut

@dr Start from C lang, end with C bots.

How come we ended up with so many Cs. Codex ChatGPT Claude Cursor Copilot ...

@samhenrigold This file is enormous. Let me do some refactors... After cleaning: 2500 lines.

yeah you wrote it dumbass

@sickdotdev Claude: human, give me the task. Make no mistakes.

when Claude already wrote like 3k lines of code and I notice an error on my prompt

@malva_0x @PranayReddy05 @dylanmatt Right, it isn't translation. But attention mechanism was originally introduced to solve translation, beating all other methods at the time. That shows attention can do this task well. Transformer also uses attention, so it's not hard to see why it can translate well.

@adyingdeath @PranayReddy05 @dylanmatt Next-token prediction isn't translation. The surface behavior can resemble it. The origin story keeps getting cited as an explanation when it's just provenance.

I asked Claude to convert a German-language PDF to plain text "so I can send it to Google Translate" and at the end he was like, "FOR THE RECORD, I can speak and translate German perfectly well myself"

@PranayReddy05 @dylanmatt Disagree The Transformer architecture behind most LLMs was originally introduced by Google for translation tasks, outperforming other methods at the time. Later, it was adapted to predict the next token, eventually enabling models to speak like humans. It's good at translating.

@dylanmatt I don't think Claude or any other models are that accurate at converting languages.

@dylanmatt Claude be like: why use that. I'm great.

@scaling01 It seems they're running into issues with a lack of computing power. Now they're starting to do something about it.

Anthropic removed Claude Code from the Pro plan I'm obviously going to cancel my subscription if I lose access to Claude Code Mythos was actually the top of the Anthropic hype cycle

@aliceisplaying Grok: No. I'm sick of this shit. Go ask ChatGPT. ChatGPT: No. I'm sick of this shit. Go ask Gemini. Gemini: No. I'm sick of this shit. Go ask Claude. Fine loop.

@parmita Maybe because it costs some tokens and Claude 4.7 wants to earn more money for Anthropic.

Um. What?

@hiarun02 Now I know how to get ChatGPT Plus for free.

Bro skipped the paywall with words.

@ns123abc What will happen when it came to 100T?

THE ROAD TO GROK SUPERINTELLIGENCE

@FlorinPop17 Building strong personal brand will bring back a huge rewards. We should keep working on it.

Focus on building a strong personal brand.