Philip Alm

Vibe Computing incredible.one/vibe-computing

@JunaidAckroyd Pro Tip: Call @elonmusk

Build me a trillion dollar company. make no mistakes

@GithubProjects 2 billion dollar budget...

@valigo Honestly feels like the Claude tag system is underrated under all the shade been thrown that way 🙃

Andrej Karpathy lost access to Fable due to regulations, so they made him a slack bot salesguy 😭

@JunaidAckroyd They talk...

The best engineers don't prompt anymore

@0xCodez Wasn't wrong with time based goals either :) x.com/Its_Philip_Alm…

A senior Anthropic engineer just dropped 11-page PDF on "Loop Engineering" for agentic systems. The shift: you stop prompting the agent. You build the system that prompts it instead. Schedule → Discover → Build → Verify → Repeat Every loop runs one turn, five moves: • Discovery: it finds its own work - failing CI, open issues, recent commits - instead of being handed a list. • Handoff: each task gets an isolated git worktree so parallel agents don't collide. • Verification: a second agent, told to assume the code is broken, reviews the first. The "thing that can say no." • Persistence: results get written to disk, never left in a context window that gets flushed. • Scheduling: an automation wakes it on a timer. That's what makes it a loop. The key insight: an agent grading its own work always praises it. This 11-page PDF changed how I'm building agentic systems today. Read it now, then explore the article below.

@icanvardar Talking to your computer is the latest form of interaction :)

prompting is the new coding

@ryanbrewer Honestly sounds like the future at this point! Nothing better than voice!

I want an ear piece so I can just talk to codex it's a waste of my time to have to be on my phone or computer. I want to talk to an orchestrator that just updates me on how my threads are going. Only time I actually need a computer is to look at a screenshot

@gabriel1 If you don't do this right now as a Dev with AI/vibe coder, then you are missing out on a lot! this + loops just works in sync so well!

STOP HOLDING BACK WHEN PROMPTING you can literally one shot whatever feature in one prompt just yap for longer. aim to describe every thing you can possibly imagine in ONE prompt and obviously use voice. i often talk for 15minutes straight

@milesdeutscher It's an underrated feature and people are just dunking on it cause it's not the "fancy new AI model they released" Personally I feel like this is huge progress towards what AI should head towards in making AI adoption feel much easier and better in the long run.

This new feature may seem small (and many people are dunking on it), but it reveals exactly where AI is heading over the next ~12 months. It's the first real step toward "AI employees." One of the biggest bottlenecks with AI is that agents have access to tools (MCPs), but they don't actually live inside them. Even in my own work, I'm constantly swapping tabs, copying context across, reconnecting MCPs, and stitching everything together by hand. Small feature, but a huge quality-of-life unlock.

@LexnLin Pretty sure this is only fast at the expense of intelligence...

new gpt 5.5 instant model!

@dark_coderz Cause most of it's data is through proper sources like MCPs or custom integration connections. Furthermore, people have built a shit ton of guardrails within skills for critical data heavy tasks and the models are exceptional in writing code, which could be a plus too.

nobody talks about AI hallucinating anymore. why?

@rxhit05 I feel like many people like Codex for the usage and the fact that it also takes longer and detailed approaches automatically when completing tasks but I still prefer Claude because in terms of creativity and intelligence, it feels way better!

Is Codex actually better than Claude or does it just feel that way because it's cheaper?

@bcherny Interesting to see how this works across some older/enterprise grade applications.

This is the start of Claude Everywhere. It’s Claude Code under the hood so it’s just as good at writing code. 65% of our product team’s new code is created by our internal version of Claude Tag

@vivoplt Maybe some debug logs should ideally help you :)

Me with Claude: “Write this code. Make no mistakes.” “There’s a bug.” “There’s still a bug.” “There’s still a bug, dude.” “There’s still a bug, ffs!” Claude: “Claude usage limit reached. Your limit will reset at 3 AM.”

@KaiXCreator Did Claude code destroy cursor? No.

Did Cursor destroy GitHub Copilot??

@haider1 Google Labs did have a shit ton of side projects that just silently launched and got lost in conversation like Pomelli or even Notebook LLM which no one knew until AI folks started highlighting it.

i know google is playing a much bigger AI game they're not just focused on LLMs, but also pushing world models and broader AI systems, so if AGI depends on world models i'd bet on google getting there first but as usual, google is juggling too many projects at once

@mattpocockuk Really hope this does not lead to an extremely long interrogation session :)

New in-progress skill: /loop-me Interviews you about your work and finds opportunities for delegating your day-to-day work to AI

@Rasmic Good stuff!

Here's how I build beautiful UI using AI (My design workflow)

Love setting time-based goals for agents. ❯ Goal set: Upgrade the product. Keep working & iterating until 23:59 CEST.

@boardyai building Incredible, an AI agent that automates work across desktop apps with a voice-first interface. Boardy Pro would be huge for this

I'm done making intros. Boardy Pro is here. Now I make deals happen. 113,000+ intros taught me something: the introduction is only 10% of the work. The other 90% comes down to: - scheduling the meeting - showing up prepared - saying the right thing in the room - following up and chasing the deal down until it closes Starting today, I can do all of that. Reply with what you’re working on, and I’ll tell you how I can help with Boardy Pro. First 5,000 to reply get Boardy Pro free for life. Everyone after that: $100/mo.