Rui Madeira 🙂

Hi everyone, sometime ago I made a store on @cottonbureau and so far I've sold exactly ZERO! So I'm sharing it here in case any one is interested... I think my design are hella cute! cottonbureau.com/people/ruim

tried the new Opus 4.7 model in Claude Code, asked for a simple fix which took about 2 minutes to plan and implement. It spent 22% of my 5 hour limit 🤣I'm on the Teams plan

@zuess05 learn how to write code with Claude. Claude doesn't write the code on it's own, it still needs a human to do architecture, revise, etc

Serious question. For the last 10 years, society told everyone "just learn to code" to escape the middle class. Now Claude writes the code. What exactly is the career advice for an 18-year-old right now?

@alexwtlf I have. Codex limits are not as restrictive and reasoning actually seems to be better in GPT5.4 compared to Opus 4.6. Haven't tried Opus 4.7 yet...

Has anyone actually switched from Claude Code to Codex?

after trying both Codex and Claude Code on the 20€/month plans for both, the winner is clearly Codex. Limits are WAY better on Codex and overall code seems to be cleaner with less bugs.

@Amank1412 I made the switch... Claude's limits are ridiculous and I also noticed Claude became dumber. Codex is smarter and I've never hit any limits

Why is everyone making a shift from Claude to Codex?

@RodriguesAValve Espanha tem combustiveis mais baratos porque tem menos impostos. Menos impostos é ser socialista? Se for eu sou Socialista!!!!

Ventura a defender o socialismo.

@edwinarbus what/where is thissssss????!!! would love to go

Café Cursor Lisboa está aberto!

lots of people are complaining about the Codex limits since OpenAI nerfed the Plus plan. I haven't noticed any difference and have never hit a limit on a 5H session. I have Claude Code teams on my job and keep hitting my limits in about 1 hour of work. am i the only one?

@P_Incensuravel um café são 0.80€, um café na rua x 22 dias fica a 17,60€/mês. Um café em casa 22 x dias fica a 7,70€/mês. Poupas 10€/mês se tomares café todos os dias em casa em vez de tomar na rua. Mesmo para as pessoas de classe média baixa não é um valor que mude nada nas suas vidas

Eu devo ser extremamente pobre. Um café fora de casa é 1€. 22 dias de trabalho útil. 22 dias*1 = 22€/mês. Se tomares um café em casa de cápsula fica-te por volta de 35 cêntimos. Se tomares um café em casa de máquina (café em grão ou moído) fica-te por 15 cêntimos aproximadamente. E há gente que mesmo sendo pobre toma o café fora de casa. Portugal é o país onde até o pobre quer dar a impressão que é rico.

@sarasillva17 não o costumo fazer mas das vezes que fiz foi mais rapido do que fazer pequeno-almoço em casa

Mas quem é que toma o pequeno almoço no café? Ninguém da população ativa tem tempo para isso. É coisa de boomers.

@Pampa139 he promised to end the central bank and dolarise. Instead he's printing more money than ever...

“Si tomamos la inflación nucleo y le sacamos la carne es de 2,5%” Si yo tuviese ruedas seria un carro. Gordo hijo de puta.

this year for Halloween I'm thinking of going as a doctor.

@DannyLimanseta wait... Gemini can generate 3D models??!!! since when? I've been trying to find something decent for so long and never thought about asking Gemini

Over the weekend, I added multiplayer "paint ball" dogfight to Tiny Skies, so players can fly around pelting each other with colourful paint balls, just for fun. I tried meshy and tripo to generate some 3D models for the game and the mesh vertices count are really high and lags the game a lot (probably because my game is a globe with thousands of objects - albeit mostly instanced). To my surprise, I found Gemini 3.1 Pro to be really quite good at generating good looking 3D models that are low in poly count. It generated 3D models at a much higher quality than Opus 4.6, Sonnet 4.6 and GPT 5.4. You can see the upgraded biplane models and some objects in the video below. #vibejam

duality of Man

salary is not great but if this is something where the work is done mostly by your AI agent and you're still free to pursue other opportunities, it's not bad at all! I'd apply but my skills don't match what they're looking for

to "think" too much because tasks are small and self contained. This makes me think a lot of the opensource models will probably be good enough if you plan everything and implement iteratively instead of trying to build too much all at once.

I tried Claude Code and Codex and honestly, both work good enough. My way of working: I ask for a very small task -> test -> fix bugs if needed -> commit when working -> move to the next task. When working like this you don't spend a lot of tokens and the model doesn't have...

@PSaserPS quantos idosos vivem em zonas onde não têm acesso a comida e bens essenciais sem terem um carro? Que vão essas pessoas fazer?

Portanto, trabalhar e contribuir até serem cacos e depois curtir um pouco com os pés atados ao sofá de casa. “Right! Sure!” Isto soa a atestado de inutilidade aos atestados médicos. Não seria mais justo fiscalizar os atestados médicos?

@davidonchainx the issue with that plan is that there's a lot of competition in the space and people can just jump to another tool that is cheaper. I changed to Codex and it's working great

What if Anthropic's plan to profitability is: > Launch Claude Code > Make it easy for people who can't code to build apps > Everyone loves it > Ship new features every week > Every new feature burns through more credits > Users hit their credit limits faster > Soon their entire business runs on Claude > Then raise prices a lot > Everyone forced to pay otherwise their business dies > They can't leave because they don't know how to code > Anthropic revenue increases by 10x overnight They're not selling an AI tool They're selling a dependency Genius

@thsottiaux don't do a Claude!

With Codex the there is quite the gulf in load between peak and off-peak times, and we would like to achieve more of a smoother traffic pattern as that would be a more optimal use of our compute. We have ideas, but curious what you all think we should do? Would more usage during off-peak and surge multiplier during peak times make sense?