Jorge Mena

3 years ago I was on minimum wage in Chile. I taught myself to code at 30 → became a Senior Engineer in the UK (x15 income). Now I’m building Andy: an AI agent that answers customers + captures leads 24/7 on your site & others channels. Try it free 👇 andypartner.com

@alexandr_wang @grok price?

the muse spark API will be coming soon! we have been thrilled with the amount of excitement amongst developers who want to try muse spark inside their agentic harnesses stay tuned!

@Isaac27330420 @midudev Hay muchas enterprise que lo ocupan. Yo ocupo codex y cc personalmente y por medio de donde trabajo ocupamos GitHub Copilot y para ser sincero, ah evolucionado muchísimo. No está tan atrás de cc o codex.

@midudev A quien le importa GitHub copilot sin ánimos de ofender a nadie aquien esta muerto eso

Microsoft acaba de lanzar un curso gratuito y de código abierto: GitHub Copilot CLI para principiantes. 8 capítulos con proyecto práctico, usar el contexto, crear agentes personalizados, trabajar con skills, conectar servidores MCP y mucho más. → github.com/github/copilot…

Nah you’re not quite right. Codex 100% of the times. I’ve literally ask to plan tasks so many times to outputs and codex and the ask both to choose what’s the best one. Literally 100% times both choose codex plan. It just doesn’t make sense to use Optus, you’ll get bugs and bad implementation all the time

I’ve used Opus 4.6 and GPT 5.4 on a mix of projects since release, and want to break down where I think they uniquely excel. It’s more nuanced than you’d think! Rigor of code - GPT 5.4. It goes the distance validating its work without asking. Opus needs explicit instruction to do this, and even then, it misses more edge cases. Clarity of code - Opus 4.6. Claude is a better communicator, which carries into the code. Variable names are clearer and less mechanical, which improves reviewability. This is very important since code review is the bottleneck for most engineering teams. It also adds the right amount of doc comments. GPT simply never comments or explains its work; it’s like working with an obtuse engineer that wants the solution to speak for itself. Sometimes it does, other times not. Similarly, rigor of plans goes to GPT 5.4, while clarity of plans goes to Opus 4.6. An interesting point though: GPT performs better talking through a strategy without a plan, while Opus needs planning mode to put in any rigor. I find myself forgetting plan mode altogether using GPT 5.4. Quality of research - toss-up. Opus spends longer researching with web search, but GPT spends longer studying the existing codebase. You may think codebase research matters more, but researching how others solve the same problem can be just as important. Maybe more important for greenfield. Quality of conversation - Opus 4.6. It’s just better to talk to, which matters using these things everyday. GPT 5.4 was clearly trained to challenge the user more, which results in a tendency to *always* say you are wrong. I’ve had bizarre interactions where GPT claims something is “not quite right,” the restates exactly what we’ve decided on in the last turn. On a personal level, it’s annoying. On a practical level, it makes iteration on a plan slower. THAT SAID, it takes sufficient pushing for Opus to challenge your thinking in this way. Simply say “I’m impartial” and ask questions to avoid that, as you would a person. Overall winner - Opus to make it work, GPT to make it good. I don’t have a good system of when to switch tools, but on average, I prefer Opus early on and GPT for optimization and discussing architectural decisions. Opus is also better for any design related tasks (but state management in frontend apps is better handled by GPT).

The smartest people are all self-taught, even if they went to school.

@developedbyed we all know is not good for frontend brother, not surprise here :/

Opus 4.6 vs GPT 5.4 (high) (2/9) prompt: Build a premium Vite + React landing page for Aether, an AI creative studio, with dark-mode design, high-end animations (Framer Motion), and conversion-focused sections (hero, features, demo, pricing, testimonials). Not sure what GPT 5.4 was going for here with the busy Headers, ugly blurs...Opus 4.6 is a winner here, since this is a oneshot I'd much rather start with the left one. (gemini 3.1 pro also performed better on this test) Still early to tell how well it holds up with design, way more tests left to do.

@robj3d3 try to have a baby or two, then come back and talk to me xd

If you’re a founder you should be training like an athlete: > 8hrs sleep > no alcohol > daily sun > lift 4x a week > 10k steps/day The #1 reason founders fail is because they quit. If your body is fit, you won’t quit.

@thsottiaux frontend

With GPT-5.4 out. What should Codex ship or improve next?

It’s not about junior vs senior, it’s about “good with AI” vs “not good with AI.”

We have raised a $110 billion round of funding from Amazon, NVIDIA, and SoftBank. We are grateful for the support from our partners, and have a lot of work to do to bring you the tools you deserve.

@hugorcd Big inspiration indeed 🙌, amazing work! cheers

@JorgeMenaDev Not at all 😂 For me, the purpose of this kind of template is more to provide best practices, inspiration, ideas, etc.!

Casually killing my SaaS 😂

@HsanC_ I've used claude code since day 1, months on pro max, and i can tell you, codex is so much better, is not even close

I don't understand why people still use Codex. Claude is at least 2x better IMHO.

@antonyemholland @HsanC_ thought the same hahaha

@HsanC_ No way man this is rage bait surely.

@OpenAIDevs London?

Looking for a Codex meetup in your city? Our ambassador community is bringing Codex to you. Create and ship projects with your local developer community, compare workflows, grab coffee, and meet people building with Codex. developers.openai.com/codex/communit…

@Jicecream Yeah… however, it makes me think, has ever a good thing come free? 😂

@JorgeMenaDev If only it didnt make me feel shit the NEXT day...

What's something you bought that actually made you happer?

The fact your agent can fetch logs from your services in production is insane .e.g vercel or your db in convex. Every company should try to aim to do this either by mcps or skills. What someone takes to debug probably hours could be done in 2 minutes if you're using ai properly.

Careers are dead. Jobs are dying. Opportunities arising.

@acredix/web scored 88/100 (Great) on React Doctor. Run it on yours: react.doctor/share?p=%40acr…

I've been reading more and more about creating software for agents, no for human, this is a great example 👇 your splitwise is dead in six months. here's why. think about the last time you split a bill with friends. someone takes a pic of the receipt. everyone does the math. someone remembers they owed money from last month. what should take 30 seconds turns into a 10-minute negotiation. within months, that problem vanishes. you pay. your agent talks to their agents. you get a message: "steak and coke zero again? approve $42 to jon?" you tap once. done. this isnt really about bill splitting. its about what happens when software stops talking to humans and starts talking to other software. right now, great software means great ui. beautiful buttons. smooth animations. hell, we obsess over onboarding flows. but if agents become the primary user, none of that matters. an agent doesnt care if your interface is pretty. it only cares if it can access your api. if an agent cant call it, compose it, and automate it, your software becomes invisible. suddenly the infrastructure layer, the composability, the api design—thats everything. the surface becomes secondary. we're watching a fundamental architectural shift happen in real time. and most products arent ready. the inflection point is real. ai writes 99% of the code being shipped today. not in labs. in production. yesterday i planned features. today agents build them overnight while i hit the gym. this isnt hype. this is already happening. if your job involves repetitive work, you have two paths: automate it yourself or watch someone else do it and make you obsolete. either way, the automation is coming. here's what actually matters now: picking the right problems to solve. intelligence is on tap. software is free. the scarcity isnt compute anymore. its judgment. taste. knowing which problems are actually worth solving. that blank canvas problem is real. yes, you can build anything. but what should you build? thats the hard part now. distribution, brand, trust, expertise—these become the real moats. if youre building anything, build it for agents, not humans. make it api-native. make it composable. everything else is secondary.