Deniss Alimovs

In March we wrote one line into our strategy: we are not competing to own the chat interface, we are competing to own the context layer beneath agentic work. On June 30 we deleted our in-app AI chat. Months of work, gone in one commit. Ten days later, Claude's desktop app shipped an in-app browser. The chat we refused to keep building can now show our canvas right beside the conversation. Strategy is deliberately saying what you are not going to do. Writing it down is what makes the deletion cheap.

@jasonzhou1993 Love the new microphone 🎤

I gave my coding agent a map across 3 of my repos: - Reduce ~50% token - a grep hook so every grep call has much richer info - trace any function's call chain + blast radius Really can feel my agent is getting A LOT smarter; I added to my codebase setup skill + 8-min deep dive: github.com/AI-Builder-Clu… Copy it. Adapt it to your codebase.

@jomatech 💯

@denzoADHD Wait, that should be my actual “app” then.

I'm building an app to make $10,000/Month and I'll record everything... I want to document my WHOLE journey building an app, marketing it, and making money off of it. Is it really that easy? How hard is it? I've been seeing so many X users online talking about how easy it is to make money online with AI now. I want to test it and show people the reality of making a business online.

@japborst @karrisaarinen @linear I am predicting work without tickets coming soon. Instead, history ledger of collaboration on improving goals.

@karrisaarinen @linear Serious question: if I can create a linear ticket to spin up Claude to open a PR, why not go straight to Claude? There's ticket history, but so is PR history.

The software factory is humming along in @linear. An issue comes in from Intercom to Triage. Linear routes and categorizes it, investigates, opens a coding session, and has a fix ready for review 10 minutes later. We’re now merging 50 to 70 Linear agent generated fixes per week just from Triage. (Usage is priced with clear token-based credits. No monthly AI fees, no expiring credits.)

@hnshah Why do you think that "It cannot tell you which customer problem is worth solving, which trade-off matters, why a feature will fail"?

@clairevo It's pretty wild ha! How do you keep track of all the moving parts and how do they share context?

i am at a very confusing time in my life. cursor is writing my newsletter. codex on rev ops. claude teaching summer school. devin still writes code. bugbot acting like its the only staff engineer worth a damn. 17 subagents doing god knows what. i think i'm a manager again?

@trackmatte @ericosiu What is your team building with ollama? Seems like day and night in capability between ollama and Opus 4.8

@denzoADHD @ericosiu They will to lock you to their model. I’m running this now in my coolant using OpenClaw in slack. Free main agent ollama. Subagent specialists. My team is on fire building skills and tools. Super powers.

In twelve months, EVERY company will be running a Company Brain. The teams who build it this year will spend the next year compounding. Everyone else is going to play catch up. Here's what it actually is. You connect your Slack, your GitHub, HubSpot, all your tools into one intelligence layer, then build the org chart around it: a main brain up top, a fleet commander running the agent fleet, specialist sub-agents handling execution. The reason it works is change management basically disappears. Your team already lives in Slack. You're just adding agents to the room they're already in. You NEED to start building yours now. In a year this will stop being an advantage and will become table stakes.

@pratos_ Why do you think? As how would you monetise decision traces?

Decision traces are very likely to be the goldmine. Excellent article.

I keep thinking about this random thing from when I was growing up in Latvia. There was this place in Riga called LIDO. Not just a normal restaurant. More like this massive food destination. The big one on Krasta Street. As a kid, it felt like food heaven. You would walk in and it was this huge, warm, wooden, almost market-style place with chefs cooking everywhere. Fresh food being made right in front of you. Proper Latvian food. Homemade-feeling food. Soups, meat, potatoes, salads, desserts, everything. You'd grab a tray, walk around, point at what looked good, and somehow everything tasted amazing. It was the kind of place where everyone went, so of course the car park was massive. And at the entrance to the car park, coming in from the road, there was always this one guy standing there like a traffic controller. But the funny part is that there was only one way to go. That was it. Cars would queue up, slowly roll forward, and this guy would just stand there pointing in the only possible direction. And I remember being in the car as a kid, being such a little shit about it. Like, "What is this guy even doing? What are the other options? Why is he pointing? We can only go that way. Get a real job." Just pure childhood arrogance. Zero empathy. Full confidence. And now I honestly think I've been cursed for it. Because I swear this is exactly where white collar work is going. A customer has a problem. An email comes in. AI reads it. AI turns it into insight. AI shapes the solution. AI writes the code. AI tests it. AI ships it. AI writes the marketing. AI updates the docs. And then there's me. Standing at the entrance. Pointing. "Yes, customer problem. This way please." That's the job. The future of knowledge work is me becoming the LIDO car park guy I mocked as a child. This is karma. And honestly, fair enough.

@zachtratar @TylerMaran @notion Nice! I wonder if it is possible to track back and figure out how it slipped for a case study. Like, as in, did the human make a mistake during the input by not specifying a way to validate completion or the agent's inference halucinated

@denzoADHD @TylerMaran @notion We use a bunch of models so it depends on the team that owns sidebar/onboarding. Not sure. I do know we have the Codex bug review bot automatically reviewing PRs.

about to churn our team from @notion because I can't for the life of me figure out how to get rid of this garbage

@zachtratar @TylerMaran @notion Honest question. if you are willing to share, what model was it coded by and was there another model, if so same or different model type that tested it?

@TylerMaran @notion We are working on this now! Apologies.

Writing notes faster just gives you a bigger pile you cannot trust. Capture is the easy part. Keeping it true as it grows is the work most tools leave to you.

Note piles don't die in a fire. They die quietly. A note saved and left unlinked. A claim with no source. A fact that stopped being true in March, with nothing to catch it. Six months in, your knowledge base is a search over a junk drawer.

@getheroik @zachtratar @kevinrose @craftdocs In creatornotes.app every card on the canvas is a database record: typed, versioned notes, with edges as typed relationships. Click any card for the full markdown doc. Happy to walk you through it.

@zachtratar @kevinrose @craftdocs I want objects on the whiteboard to correlate to records in the database. So the whiteboard becomes visual interaction w records in a database w clickable depth. Data vis to explore as well.

Notion feels bloated these days, Obsidian is great but I don't have time for all the configuration/extensions. Been loving the latest from @craftdocs - less complexity but native BYO AI keys + mcp support.

@kevinrose @craftdocs Is it for solo/personal notes which are occasionally shared or are you trying to create a company brain that scales for teams?