Fedor Korotkov

New York is about to make a massive mistake. The NY State Senate is advancing a proposal to decouple from federal QSBS (Section 1202) — the tax provision that lets startup founders exclude gains on qualifying exits. If this passes, founders would owe 10-13% in combined state and city tax on exits that are tax-free at the federal level and in nearly every other major tech state. Even worse: it's retroactive to January 1, 2025. This comes right as the federal government just expanded QSBS benefits and New Jersey moved to full conformity. New York wants to go in the opposite direction. As a seed investor in NYC who has backed hundreds of companies, I can tell you: founders are mobile. If New York becomes one of the most punitive states for startup exits, the best founders will simply build somewhere else — and the jobs, tax revenue, and innovation will follow. NYC has built something special over the last two decades. This proposal puts it all at risk for a short-sighted revenue grab. If you're a founder, investor, or anyone who cares about the NYC tech ecosystem — please sign the TechNYC open letter before Monday below 👇🏾👇🏾👇🏾 Keep building, NYC 🗽

Introducing mtell -- a simple CLI tool that can visually control any computer over VNC 🎉 github.com/cirruslabs/mte… Type text, press keys, wait for text on screen, click buttons via OCR, or hand off complex UI flows to @OpenAI Computer Use. Here is a demo that uses a Tart VM 🍏

@mir_nomer_nol For a decent model you probably want a Mac Studio but I haven’t played too much with local LLM besides gpt-oss ones. I was thinking about use case with a subscription for which people buy the cheapest $600 Mini with M4 and 16GB of memory.

@fedor Mac mini is good for running local llms. Not sure how it will work with A-generation chips

Is MacBook Neo a better replacement for OpenClaw than Mac Mini? Builtin screen, keyboard and portability is more important than performance if you don’t do heavy coding.

@jetserge @intellijidea Need to do a CA reunion!

@fedor @intellijidea I was there too, and I felt the same way. Fun times! Those I/O business trips likely shaped my decision to move to California in 2017.

Just watched @intellijidea documentary and it brought a lot of memories and nostalgic feelings. One of the highlights for me during my time at JetBrains was attending Google I/O 2013 when Android Studio was announced. I remember feeling of keeping this secret and seating during the keynote. The room just exploded during the announcement! youtu.be/Kourq_Lz03U?si… I was a student working full time back then. In a month after the conference I got my H1B visa and moved to San Francisco later that year.

I still have that orange t-shirt!

Lobsters at #ClawCon NYC. Only saw such fancy food on meetups around 2013-2014 golden era. Energy also kinda matches.

Chrome has been crashing for me for months. I tried to investigate. No crash reports, no logs in any location, stripped down all the extensions. All! Still crashing and even more frequently. Switched to Safari. Wish me luck.

@auchenberg Holly molly. I though I was fancy with my Breville Dual Boiler. 😅

Bliss. ☕️👌✨

@marcaruel Oh yes, they always have issues with macOS runners specificly. If you ever plan to self-host them for your projects please keep in mind @cirrus_labs. We perfected macOS infrastructure.

@fedor GitHub hosted runners had really bad uptime in the recent weeks and in particular macOS ones. The tweet above is that I suspect the increase in coding agent generated code created a spike in usage and their fleet struggles to survive the demand.

that probably explains why GitHub macos runners are dead

@auchenberg Very impressive even for an 8B model. Haven’t found information how it scales with a model size. Definitely worth keeping an eye. 👀

🤯

@jarredsumner I have a skill with bunch of GraphQL to do exactly that. Kind of ugly but GraphQL API is really powerful.

feature request for GitHub’s gh cli: a subcommand for agents monitoring PRs - view unresolved pr review comments as markdown or xml with file:line - show failing gh action logs, maybe filtered to near the error - show lint errors from gh actions

We just made the backbone of our caching infrastructure public under @fairsrc. It handles billions of cache operations for our CI services — and now it can help you, no matter which CI system you use. Introducing Omni Cache — your friendly neighborhood sidecar for all CI caching needs. It exposes multiple cache protocols on a local endpoint while storing blobs in S3-compatible storage. Instead of funneling traffic through a centralized cache proxy that becomes a bottleneck at scale, each CI job talks directly to S3 through a local sidecar. No extra ingress/egress fees. github.com/cirruslabs/omn…

@anthonyronning @samhogan @kylegalbraith @jefftangx @useblacksmith @depotdev @rwx_cloud Checkout @cirrus_labs. We have the leading macOS infra at a fixed monthly price for unlimited minutes.

@samhogan @kylegalbraith @jefftangx @useblacksmith @depotdev @rwx_cloud Well check them out. I tried blacksmith and saw worse performance with 1:1 image specs and was also missing critical macos infra.

Who are you guys using for GitHub Actions alternative? It's been down today and in general slow/bad UX I've seen @useblacksmith @depotdev @rwx_cloud

Why no one is talking about the Plan Mode in Codex App? It was the only feature I was missing.

@mir_nomer_nol Glad to keep it! 😅

@fedor Welcome back! You really have nice acc name from those early days, great that you saved it!

Restored my account after two and a half years! Was suspended for alleged spamming right around Twitter/X transition and then cutting API access and tightening rate limits. My bet is on false positive due to my cool username and constant spam in mentions.

IMO looks like sheltering all the shenanigan Twitter/X/xAI has been doing since Elon took over. On a bright side, after an IPO we’ll get a little glimpse inside. But now I’m not planning to invest in SpaceX shares personally.