all the nuggets that cursor’s been dropping seems like the next evolution of cursor will be to allow users to build full end-to-end applications with all the services needed in one go
not a prompt-by-prompt vibe coding platform like what we’ve got in the market. but something that takes 10-20+ hours to run and no human intervention
We built a browser with GPT-5.2 in Cursor. It ran uninterrupted for one week.
It's 3M+ lines of code across thousands of files. The rendering engine is from-scratch in Rust with HTML parsing, CSS cascade, layout, text shaping, paint, and a custom JS VM.
It *kind of* works! It still has issues and is of course very far from Webkit/Chromium parity, but we were astonished that simple websites render quickly and largely correctly.
despite all these models becoming so smart and surpassing every known benchmark known to man, it’s crazy how differently they all behave no matter the guardrails
all routes lead back to eval
with agents getting more sophisticated and deeply integrated into every aspect of a business, every product needs to prioritize data ontologies
for the longest time, data ontologies felt complicated and limited to data companies. but it's been a major unlock for agents to understand complex workflows and not limited to specific context
i just interviewed a 100-year-old with chatgpt voice
she lives in the mountains of Sri Lanka at 6000ft of elevation, owns a farm, and has a simple beautiful lifestyle.
we chat about health, family, living a fulfilling life, and more!
couldn't be more proud to announce our new product, Cheers GEO. We help local service businesses (plumbers, electricians, etc.) get recommended on ChatGPT and other LLMs.
a lot of you asking if i regret open-sourcing Stack Auth after this whole thing.
hell no. i actually hope MORE people copy us.
they can fork the entire repo and steal every feature. they can try to compete. but here's the thing: we'll still win.
they're spending their time parsing git commits trying to understand what we just shipped. by that time we're already three features ahead.
they're debugging code they didn't write. we can ship 100x faster because we know exactly why everything works the way it does.
let them copy, in the end it's free marketing for us. their customers will eventually discover that we are the original, and when Authfy struggles to fix their scaling issues, we're the ones they reach out to.
the codebase is only 10% of our company. the other 90% is speed of execution. and if all you do is copy code, you'll never be *pushing* the boundary of what's possible.
so yea. this was bound to happen at some point. no regrets!
(although, i do ask that our copycats respect our open-source license, which afaict Authfy did not. oh well.)
claude code needs a /improve command
- goes into plan mode or uses similar logic
- deconstructs existing prompt into more details
- asks questions
- provides a new and updated prompt
- adds new improve logic to memory and continuously updates
unless this already exists?
in person interactions are going to matter more than ever
maybe tech will go the opposite direction (already is), but for the rest of the world will push for in person meetings and conferences. we've experienced 100x better results from in person discussions and demos
incredible launch and so excited for this project!
one of many issues in Toronto: unnecessary spending. i.e. don't spend millions to rename Dundas Square based on vibes
🚀 Announcing Build Toronto
Toronto is Canada’s largest city, its economic engine, and its greatest opportunity. If Toronto thrives, Canada thrives. But we all know the city faces deep challenges: unaffordable housing, strained infrastructure, governance gridlock, and a lack of urgency.
That’s why Build Canada is proud to launch Build Toronto – the first municipal project of Build Canada – to raise the level of debate and spotlight bold, practical ideas that can move this city forward.
We are equally proud to welcome @ericdlombardi as Chair. Eric is a civic leader and housing advocate whose work with More Neighbours Toronto has made him one of the city’s strongest voices for change. He will help guide Build Toronto as we put forward ideas that support growth, prosperity, and ambition for Toronto’s future.
Over the weeks ahead, Build Toronto will publish frequent memos from entrepreneurs and civic leaders on Toronto’s biggest challenges and opportunities. From housing and transit to governance and economic growth, these memos are meant to push all of us – citizens and leaders alike – to think bigger about what Toronto can be.
Toronto has the talent, energy, and openness to lead. What we’ve been missing is urgency. Build Toronto is here to help change that.
👇 Sign up for updates on our website