Flo @marginnoteslog

@thsottiaux Announce beforehand if you are going to remove a model. I just noticed 5.3-codex is gone, and I haven't seen any info about it.

And now its gone Error: Codex error: {"type":"error","status":400,"err or":{"type":"invalid_request_erro r","message":"The 'gpt-5.3-codex' model is not supported when using Codex with a ChatGPT account."}}

5.3-codex is my favorite coding model so far. Hits the sweet spot for price / intelligence — 5.4 and 5.5 are too expensive for the delta.

The fallacy of this is that more creates more. More hours, more hiring, more something. And it is true in a sense. If you put in more work, more work will happen. But I think for most startups, the leverage is really in how differently you approach the problem, how well you cultivate your team, and the strategy. Any large company can outspend you on hours. They have thousands or tens of thousands more people, spending more hours. If hours worked were the metric, every large company and government organization would always win and do the best work. More hours, better output. This thinking is often representative of younger founders, where the startup becomes their identity and life. They have a hard time doing anything else, and cannot understand that your work is not the person that is you. But activities outside of work can grow you as a person too and make you do better work. I’ve never worked this way. As a designer, I always saw the need to take a step back, to take a break. At times, I might work 12 hours or 16 hours, or whatever amount was needed, but it wasn’t the norm. You just can't grind design, you need inspiration. But taking that step away from the work, would give me more perspective, inspiration and I could approach the problem differently or I could just see the solution. Grinding is never good for any creative problem, and startups or creating new products are often mostly about creative problem solving. Grinding works ok for email jobs, or where you just executing on very clear playbook. With Linear, we’ve never worked this way. We work reasonable hours, 5 days a week. All of us founders have families. Many of our employees have families. I personally stop every evening, spend time with the family, cook dinner for the family, eat dinner together, and focus on things outside of work. Sometimes I work in the late evenings or weekends, but to me the pride is that I don’t need to. Company should be succesful without it. My goal is to build a company that is sustainable in the long term, and doesn’t require heroics or personal sacrifices every single day. There are times when our team is heroic. Launches, incidents, some other work that just needs to be done. They will work late into the night because they know it is the right thing. But we don’t require that every day or every week, and the more this happens, the more I think it is a failure of our company and leadership. The team and the leaders should always keep a reserve to use when something is needed. Our thinking was also that quality, which we value, doesn’t emerge from working more or stressing people more. It emerges when you create the conditions for it to emerge. Often it is the appreciation, space, time, and how the person feels. A person who is rested will do better work. I wouldn’t attribute much of our success to working a lot. The success came from having clear thinking, ideas, and focus to do the right things. I sometimes wish we could move the culture more toward a Zen master. Real mastery is not exerting the most effort. It is achieving the outcome with the least necessary effort.

Harry Stebbings @HarryStebbings

5 days ago

"If you are not working 7 days per week, you are going to lose". Corgi Insurance is the most intense workplace culture in startups. - The company works 7 days per week. - Founder (@nico_laqua) lives and sleeps in the office. - He built a cafe in the office because there was

477 41 722 3.0M 920

@thorstenball Which ones, concretely? Genuinely curious

Codex App has regressed in performance when switching between threads - UI slow. Make it easier for me to use / monitor multiple threads at the same time (split windows?) After having given enough context I would like to switch to a smaller model for implementation and then back to a bigger model for review and planning. While this is possible the UI warns me that switching models mid conversation degrades performance.

@matter Finally!

The "Accelerate grid expansion" section reads like lobbying. OpenAI proposes public-private partnerships where taxpayers subsidize new energy infrastructure and a federal authority fast-tracks permitting for data centers. Citizens get a vague promise of "lower energy costs for households." Meanwhile in Virginia, data centers already consume 26% of the state's electricity (EPRI, 2023). Nationally, data centers use 4.4% of U.S. electricity — projected to reach 6.7-12% by 2028 (DOE/Berkeley Lab, 2025). Should the amount taxpayers invest in infrastructure which mainly benefits AI be mirrored as investment back into the public? apolloacademy.com/data-centers-s… newscenter.lbl.gov/2025/01/15/ber…

Regarding OpenAI's new industrial policy paper. They name "power and wealth becoming more concentrated" as a key risk. But then: "Broad participation in the AI economy should not depend on access to the most powerful models—it should depend on access to AI that is useful, affordable, preserves people's privacy and expands their individual agency." This leaves the door wide open for power concentration. People get a share of the new wealth, but if the most powerful models are inaccessible, their agency is limited. OpenAI is trying to start a debate but they also need to protect their business. The safety argument is real. But you can block dangerous capabilities without locking away the entire model. We already do this with guardrails. Shouldn't everyone have access to the most powerful models?

OpenAI Newsroom @OpenAINewsroom

2 months ago

Read the full ideas doc on the new Industrial Policy for the Intelligence Age: openai.com/index/industri…

64 85 829 151K 297

Read @ruima's translated piece on Moonshot AI. I tried to figure out what it means for how I work: flo.fish/flat-hierarchy…

Rui Ma @ruima

2 months ago

x.com/i/article/2039…

26 131 945 545K 2K

@badlogicgames @steipete I just hope you are not getting a takedown notice

@elvissun I prefer it, but when using it, I actually hit the weekly rate limit halfway through the week, Where previously it was enough. I’m using a mix of medium to x-high for work.

The more I use AI for coding the more I think it can write good code – but it needs to be instructed to. By talking about constraints. For time savings you probably need to work on multiple things in parallel. My setup isn’t there yet so I’m taking smaller steps building agent skills in specific areas like REST API design as I go. Once the setup is ready I’ll test agents on bigger tasks.

Enjoy listening to @steipete lots of what he says resonates with me. Particularly the effect of „[every feature] is one prompt away“ and that there is a hidden cost for every feature you add.

@thsottiaux codex-cli plan mode is great, but here's a thought: sometimes I'm like 80% confident in the plan and just want to add a note before executing. Would love a "confirm + notes" flow instead of waiting for another planning round.

@romainhuet @CFJWilliamson Switching between work trees seems unreliable, or not working as I expect it to. It doesn't seem to take effect sometimes. Switching between threads is quite laggy. Integrate existing work trees into the worktree selector.

@Beelinkofficial What open source LLMs can I reasonably run on it?

@peduarte It helped me as well, but this one looks a bit better: Lift for Mac (it's smaller than the MX Vertical).

I suspect it’s not even a time-saver: reviewing AI code requires building the same mental models I would have built while writing it, but without the creative satisfaction. I’m starting to think “doing it yourself” is the more sustainable way to stay in the zone.

I’ve realized I enjoy programming less when I rely too much on AI. It’s great for skipping boilerplate, but delegating the core logic kills the flow state that made me love the craft.