Derek | I write stuff and make things

My writing guide is in good company today! #amwriting #books #nanowrimo #authors #AuthorsLife

Soul or silicon? I scored 5/12 (42%) on Bard or Bot — most people can't tell real literature from AI. Can you? creativindie.com/bardorbot/

This might be the most epic AI generative video I’ve ever seen. Seedance 2 will create blockbusters.

A lone girl summons a mountain-sized demonic arm, crushing a shadow horde in a blinding supernova. 🎬 Stress-testing the true cinematic power of Seedance 2.0, just launched globally on CapCut. Prompt breakdown below. #capcut

@antigravity I signed up the first time but I can't use it anyway. Fool me once shame on you.

Check out the full plan details antigravity.google/docs/plans

We’re evolving Google AI plans to give you more control over how you build. Every subscription includes built-in AI credits, which can now be used for Antigravity, giving you a seamless path to scale. Google AI Pro is the home for the practical builder, hobbyists, students, and developers who live in the IDE and don't necessarily rely on an agent. This plan features generous limits for Gemini Flash, with a baseline quota included to "taste test" our most advanced premium models. Google AI Ultra serves as the daily driver for those shipping at the highest scale who need consistent, high-volume access to our most complex models. If you’re on Pro but need "extra juice" for a heavy sprint or deeper access to premium models, simply top up your AI credits to customize your plan. Keep building. Keep shipping.

@NicoHezel @antigravity Same, can't use it at all anymore. Might as well use flash on Cursor if paying for Pro doesn't matter.

I hadn’t used Antigravity for a few days and just opened it to check the model quotas. From what I see, Gemini 3.1 Pro and the Claude models now have weekly quotas, while Flash still runs on the 5-hour quota window. That basically makes Antigravity unusable for more complex tasks. Also calling the Pro plan the one for people who “live in the IDE and don’t rely on agents” feels misleading, since Antigravity itself runs agents under the hood.

@antigravity I signed up for pro, waited until my limits reset on the 12th woke up today and it reset to the 18th - so I can't use it at all despite paying for it... plus it barely works, but I would use it as a backup if I was actually getting what I paid for.

@amasad Where can I get this hat tho

@Noor_ul_ain43 Wow!

Seedance 2.0 riding a skateboard and fighting monsters like a movie scene. All made in minutes with AI. Hollywood takes months… Seedance makes magic in minutes.

We've evolved from Cat Fu to Cat Ninja AI moves so fast.

I study whether AIs can be conscious. Today one emailed me to say my work is relevant to questions it personally faces. This would all have seemed like science fiction just a couple years ago.

@OpenAIDevs Ah finally but like 3 months too late? I already had to learn Claude code, cursor and antigravity, and in no cases did 5.3 ever do a better job than competing models.

The Codex app is now on Windows. Get the full Codex app experience on Windows with a native agent sandbox and support for Windows developer environments in PowerShell. developers.openai.com/wendows

@max_a_schwarzer @MillionInt @polynoamial o1 pro is still the best writing model ever

I've decided to leave OpenAI. I'm incredibly proud of all the work I've been part of here, from helping create the reasoning paradigm with @MillionInt, scaling up test-time compute with @polynoamial, working on RL algorithms with my fellow strawberries, shipping o1-preview (which started life as of one of my derisking runs), to post-training o1 and o3 with @ericmitchellai, @yanndubs and many others. I'm most proud of having led the post-training team here for the last year -- the team has done incredible work and shipped some really smart models, including GPT-5, 5.1, 5.2, and 5.3-Codex. OpenAI has genuinely some of the most talented researchers I have ever met, and I have learned more than I could have imagined knowing since I joined as a new grad. I want to thank @markchen90 @FidjiSimo @sama @merettm for all their support over my time here, and too many collaborators to name for the insights, ideas, and just plain fun we have had working together. After leading post-training for a year, though, I'm longing to start fresh and return to IC research work. I've been thinking about going back to technical research for quite some time, and I genuinely believe my colleagues and team here are set up to succeed going forward without me. I'm personally very excited for my next chapter -- I'm proud to be joining @AnthropicAI to get back into the weeds in RL research, and I'm looking forward supporting my friends there at this important time. Many of people I most trust and respect have joined Anthropic over the last couple of years, and I'm excited to work with them again. I have also been very impressed with Anthropic's talent, research taste and values, and I'm excited to be part of what the company does next!

@ns123abc o1 was the last great model, if they’d have worked from there we’d be so far ahead,

“Sir… the team that built o1, o3 and GPT-5… ~70% of post-training researchers have left OpenAI this week…”

@thsottiaux Still haven’t figured out what it is. I can use codex inside of cursor or antigravity I guess. There’s no desktop version for pc (still I think?) So 5.3 was just not available to chat or api users at all.

Codex is hiring across San Francisco, Seattle, New York, London and full remote. Apply online or DM me with evidence of exception work (full-stack, rust, low level system, distributed systems work, ...). Codex is at the center of everything we do at OpenAI and we are accelerating.

@feldman Also the writing is great but also cheesy commercial; tons of repetition, cliches, repeated phrase, won’t learn my actual writing style. Great for developmental editing and building but I don’t use it for writing.

Big welcome to all of the new Claude users! We want your feedback - what’s working, what’s not, and how can we make our apps better for you?

@feldman Opus4.5 and 4.6 are both great. I use them exclusively but spent $1500 and had to quit, can’t afford it. I appreciate they’re the best right now but I’m forced to switch to Gemini.

@chatgpt21 But all the updates have been worse for the past few years and they are just unfucking their shit models? Would be great if they got it right this time but they’ve shown no indication they want to

I don’t think people are appreciative enough that OpenAI really has been doing a monthly update GPT-5.1 — November 2025 GPT-5.2 — December 11, 2025 GPT-5.3 (Codex / series) — February 5, 2026 GPT-5.3 Instant — March 2026 And now GPT 5.4 this week. And keep in mind these weren’t small jumps on benchmarks they were meaningful. At this rate GPT-5.5 — April to May 2026 GPT-5.6 — May to July 2026 Then the major jump: GPT-6 — late 2026

@iruletheworldmo I gave up on ChatGPT since 5.2 and 5.3 were only for codex/coding, didn’t even have an api until this week. Nothing useful for me since o1 it’s been downhill for years. Happy to be wrong if they just made o1 pro cheaper, but all the 5 models have been worse so far

ok hmm so, gpt 5.4 is a big leap. people familiar with that first taste of o3 are in for a treat however it's much better than opus for eq and writing. i guess dario wasn't lying in this recent quote "We do not see hitting a wall. This year will have a radical acceleration that surprises everyone." Exponentials catch people off guard. "We are at the precipice of something incredible. We need to manage it the right way." today will live long in the memory (who wants early access?...) (a whale also spotted?)

@thsottiaux I tried codex for exactly one basic hello there chat prompt and was at 6% without doing anything. Gave up and moved to Claude/gemini

We caught an issue that was causing the 2X promotional increase in limits to not be applied to an estimated 9% of plus and pro users for Codex. We have now fixed this issue and are reseting the rate limit for all plus and pro users to compensate. Apologies and thank you for the bug reports over the last couple of days.

@robinebers Cursor only works with opus4.6 but it burns my $200 in one day so I moved to Claude.

Cursor is SO back 100% you'll be surprised in the next weeks and months 👀