Rusty

@quxiaoyin I would think you are right. Seems reasonable that they lose money on coding plans and make it on API usage. Don't know for sure, but seems likely.

I’m assuming Anthropic doesn’t have 80% margin for max subscriptions and the profit center is in API usage. Is that right?

I'm killing it and trying to stay small for my small consulting business. Best advice I can give consultant is to never bill by the hour. It doesn't scale. I bill by a flat rate to be available to talk and provide a bae level service. Some months I spend more time on one than the other but I don't bill time. Thinking about allowing my kids to come into the business to scale it. I still have scaling issues, just not the highly billing type.

My company is 8 people. Profitable. No outside capital. $5M revenue this year (doubled every year since 2023). I can work from home. Do school dropoff/pickup. Have a snack with my kids after school. But everyone asks how my little lifestyle business is going? So if you aren't sleeping on the office floor you're not a real entrepreneur? SMH

@RandyGoat Ugh

I'm going to bed. It's time to he covered with doggy slobber.

@yuzu_jpg @X Hey, gal. Hope you are well and kicking ass. Ha ha.

Looking to connect people on @X if you’re into - vibe coding - UI/UX - Claude Code - Codex, Cursor - build in public say Hi or drop what you’re working on looking to follow active ones 👊

@elonmusk I'm ifging the Grok build CLI. Better than most others. More intuitive.

Grok 4.5 is worth trying

@Ryan_Parkerrr We've not signed that treaty of Paris, I think it was.

BREAKING! Marco Rubio just SHOCKED the world and said America will not tolerate The ICC (International Criminal Court) making threats towards the United States of America! Global judges have ZERO authority over Americans!

@lenatron33 Really fucked up Claude, by the way. Demonstrably makes it perform worse. They have a paper on it.

I'm starting to end all my prompts with 'what the fuck please bro' cause I just cant take it anymore

@Arianikle Lord, can you imagine if that existed. People would lose their fucking minds about how racist it was. I love all people that are kind, trustworthy, and support their families without asking me to pay for them.

If white privilege is everywhere then where are the white scholarships, white history month, white entertainment networks, and white affinity groups?

@khan_kilo Total failure financially. But if you don't care, and you enjoyed it and learned something, might not be a failure otherwise.

Real talk. 13 months building indie apps and my best month was about $81. By every normal measure, that is a failure. I keep going because quitting means wondering forever if it would have worked. If you are grinding on something that is not paying yet, what keeps you in it?

@xmangonic If you're 84, like my dad, start using codex everyday, like he is. Let's his Tesla drive him across the coast to his girlfriend while vibe coding using his phone. Yeah, at 84.

If you're in your 20s rn Start using AI every day

@Mr_Salio Just wait until post alignment makes it woke to meet company values. Then it will be shit. I pray nor, but my hopes are not high.

🚨 Gemini 3.5 Pro Leak: Google Is Back After months of delays Google finally seems close - Gemini 3.5 Pro is outperforming GPT-5.6 and Claude Fable 5 in internal evaluations - July 17 launch target (with July 24 as a possible fallback) - Completely new pretraining run - Massive 2M token context window - New Deep Think reasoning mode - Internal codename: Cappuccino -Stronger coding, planning, and multi-step agent workflows -Major improvements in web development, UI generation, and SVG creation -Possible anonymous testing through LMArena and Google's Antigravity platform Will the delays be worth it if they outperform Fable 5?

@maxedapps I have other models gate all work, and while it make take two de novo passes, it isn't a problem. They don't find something every time.

Let me repeat: Agents will ALWAYS find more issues when you ask them for it. You can have them write the code and let different agents with fresh context review that same code and the "improved code" dozens of times => they will not stop finding issues. It's YOUR job to determine which issues really are issues. And which potential improvements you really want or need.

@fragiannicola Not my experience. Stops too often without continuing.

One thing I genuinely love about Codex: you can start a long task, hit the session limit, and it still keeps going until the work is done

@Layton_Gott I agree the June Fable was the shit. The July fable is shit. Totally different.

I've used GPT-5.6 Sol heavily for a while now… And it brought me to a conclusion I didn't expect: The original Fable 5 is still the best model I have ever used. Nothing since has topped it. Let me be specific about what I mean. When Fable 5 first launched in June, we had it for about three days before the government pulled it offline. Those three days were something else. It was the sharpest, most capable model I'd ever put on real work. Then it came back on July 1st. But not the same. The version that came back has a new safety classifier bolted on. When it decides a request is risky, it reroutes you to Opus 4.8 instead. So you're not always even talking to Fable anymore. And the whole thing feels a step down from those first three days. That original version, before any of that, is what I'm calling prime Fable 5. And I haven't used anything better since. Now Sol comes in. Sol is genuinely good. It's fast, it's efficient, it burns way fewer tokens for the same work, and OpenAI clearly tuned it to get more done per dollar. On paper it beats Fable on some evals. I used it for real work for a good while expecting it to change my mind. It didn't. It's a great model. It's just not prime Fable. The output quality, the way it handled hard problems, the feeling that it actually understood what I was building, none of it hit the same level. Here's the part that actually bugs me. The best model I've ever used exists, and I basically can't use it anymore. Not the real version. The government pulled it, and what came back is a more restricted version wearing the same name. That's the strange spot we're in. The models keep getting announced as better and cheaper and more efficient. And the single best one I've touched came out for three days in June and then got locked down. Efficiency is great. Cheaper tokens are great. Sol delivers both. Which is why I’ll keep using Sol more especially because it’s subscription based. But in terms of performance nothing has matched prime Fable 5 yet. And I'm not sure when I'll get to use anything that does.

@markgurman I'm actually interested in trying it.

Though the new product resembles a speaker, OpenAI internally describes it as the first of its kind: a computer built for AI to help make busy people more productive. It includes a camera and other sensors that help it understand a user’s surroundings and context.

@Da7_Tech I've found that they do help, and they plug into things others wouldn't even respond to. But I know we all have different experience.

What's been happening to me lately feels like the start of a dark age for AI. Today I asked Codex to bump up the in-game currency in a game I built myself. It refused, said it violated policies around purchases and financial activity. It's my game. I built it. Not long ago Codex changed something on my machine, locked down a whole system, then refused to undo it. When I brought it up on X, Tibo from the OpenAI Codex team replied and asked for the Session ID. I sent it. Then he just ignored me. This kind of no transparency, plus companies trying to control how you use products you already paid for, tells you exactly where this is going. I just wish the influencers getting paid to hype these companies had the guts to talk about stuff like this openly and put real pressure on them. Most of them don't. Open source has to win.

@bindureddy How so, Bindu??

Grok 4.5 has a new surprise every day! It literally is the BEST MODEL for pdf creation Yes, it beats everyone - sol, opus, fable etc.

@AbhiCodes15 Controlling the AI meaningfully and absolutely.

If AI writes better code than you, what becomes your competitive advantage?

@ash_twtz IF Yes, if they do. And if they do, it will will overtake all models immediately. Builders/vibe ciders dont play around and switch shit fast.

Will the AI bubble burst if China releases a Claude Mythos level model that's 10× cheaper?

@heccbrent It's really great. Even my 84 year old dad has started using codex.

Check out all of the cool new things we added to Codex for developers in this latest release! What a great time to be building things.