Greg Ceccarelli
@gregce10
Co-Founder @specstoryai. Chief link shrink: https://tny.dev. Organizer D.C. AI Tinkerers.
Wrote a white paper about the beauty and frustration of working with software agents (Cursor, Copilot, Claude Code) for 1000s of hours. - With agents, software's primary bottleneck has moved from development speed to specification clarity. - The future points to trunk-based…
I find myself yapping at zsh occasionally mistaking it for claude
Product cloning with Lovable/Bolt/Replit = espresso shot. Quick hit, quicker crash. Building a business = aging wine. Years of patience, risk, and hoping it doesn't turn to vinegar. Not the same.
so many input tokens and dev hours are being spent building lovable.dev / bolt.new / gh spark clones. the war for the top-of-the funnel in agentic vibe coding platforms is real. in the last month... emergent.sh 0.dev…
tny.dev: A tiny, blazing-fast link shortener for the AI era. Built for agents and devs… producthunt.com/products/tny-d… via @producthunt
This is part of the narrative which I believe is wrong. I believe AI can make engineers more effective. But it has a more potent effect on enabling those who aren’t engineers today bring their ideas to life tomorrow. Doesn’t make people better at specifying what they want or…
where is all the better software if ai made everyone a developer or a 10x engineer why isn't all the software better
The most obvious AI smell in writing now is corrective reframing…. the antithesis. - “This isn’t about X. It’s about Y.” - “You think this is X. It’s actually Y.” - “Everyone sees X. The reality is Y.” It’s literally the everywhere I look and way worse than the em dash which I…
A bit of philosophy about tnydev get.specstory.com/built-to-be-ou…
We are cookin' a lot of wonderful things in our kitchen ;)
Curious how AI chat exports differ? We were 🙌. In 60 seconds, our engineer @bago2k4 compares markdown from SpecStory to native exports from GitHub Copilot and Cursor.
Really curious what < $100M ARR SaaS companies have shipped with AI. Seems like basically nothing
The new @SlackHQ API rate limits are absolutely insane: For accessing messages, you can now only make 1 request per minute, with a maximum of 15 messages 🤯 And yes this applies to ALL apps, not just public ones. We can't build internal tools on our OWN data. @Benioff this is…
apropos of nothing: this is an appreciation post for @simonw, who tries out literally all the tools, and joyfully shares (for free!) what he learns in the hope that it will be useful to other humans grateful to have someone relentlessly building with the garage door open 💗🙌
Six months ago I watched Claude burn $5 to butter toast, code I could've typed in 30 seconds. Today it opens 100% of my PRs and I haven't typed a function in weeks. every.to/source-code/ho…