Sabitlenmiş Tweet
Some yearly reflections on the business, @scholarlysoft, of what we accomplished this year.
English
Kelly Sutton
16.7K posts
Kelly Sutton
@KellySutton
Married to @eastouest. Founder at @scholarlysoft
Seattle, WA Katılım Mart 2007
1.9K Takip Edilen3.7K Takipçiler
Kelly Sutton retweetledi
Way back when, I created The Zen of GitHub to direct the taste of a growing software company. I'm still really proud of it.
Fast forward twelve years. Now it's the basis for my agent instructions (with some edits). What a wild end-run.
warpspire.com/posts/taste
English
This looks really neat.
Very glad Seattle has @aviel & the foundations folks
fndtns.org/next/
English
Kelly Sutton retweetledi
Insane safety stats from Waymo.
Compared to a human driver,
- 92% fewer pedestrian crashes with injuries
- 85% fewer cyclist crashes with injuries
- 81% fewer motorcycle crashes with injuries
English
Kelly Sutton retweetledi
Kelly Sutton retweetledi
Rails and MySQL
Sam Lambert@samlambert
GitHub's performance in 2013. The web used to be so fast.
English
@samlambert Remember when we compared Azure disk speeds with a $5 DO Droplet and it rained Azure VP emails for 5 days? Good (down) times!
English
GitHub's performance in 2013. The web used to be so fast.
English
Kelly Sutton retweetledi
Cannot wait to be on the receiving end of this as a user, opted into 20+ A/B tests at once. A slightly different app for everyone all the time.
Todd Saunders@toddsaunders
The token cost to build a production feature is now lower than the meeting cost to discuss building that feature. Let me rephrase. It is literally cheaper to build the thing and see if it works than to have a 30 minute planning meeting about whether you should build it. It’s wild when you think about it. This completely inverts how you should run a software organization. The planning layer becomes the bottleneck because the building layer is essentially free. The cost of code has dropped to essentially 0. The rational response is to eliminate planning for anything that can be tested empirically. Don’t debate whether a feature will work. Just build it in 2 hours, measure it with a group of customers, and then decide to kill or keep it. I saw a startup operating this way and their build velocity is up 20x. Decision quality is up because every decision is informed by a real prototype, not a slide deck and an expensive meeting. We went from “move fast and break things” to “move fast and build everything.” The planning industrial complex is dead. Thank god.
English
Kelly Sutton retweetledi
Please drop a star on my new open source agent framework called MVidea
Itamar Golan 🤓@ItakGol
OpenAI acquired OpenClaw FaceBook acquired MoltBook Choose your company name carefully…
English
Kelly Sutton retweetledi
I don’t want to watch the McDonald’s CEO eat a burger. I want to watch Satya use Copilot
English
Kelly Sutton retweetledi
My preferred definition of "ARR" is your single-highest grossing minute of the year, times 525,600
English
@holman Bummer. Any ideas what explains the near constant outages?
English
@KellySutton been a decade since anyone with real product ideas ran the org. combine that with a large org + old microsofties wanting to cosplay as startup people, this is what you get
English
honestly the github disasterclass is pretty remarkable these days
talk about missing the boat completely, and then you put holes in your own boat and then pump more water into your boat for good measure
English
*taps the Torment Nexus sign*
Nikita Bier@nikitabier
I feel like this startup idea is fundable now.
English
Put together a blog post organizing my thoughts on AI in 2026. Hope it helps!
Some highlights:
- MCP is dumb but it's here to stay
- I don't think we’ve seen all of the creative destruction in white collar work yet
- I wish ChatGPT/Claude worked async
kellysutton.com/2026/03/04/ai-…
English
Kelly Sutton retweetledi
it is fair to say llms are arguably a giant compression of what humanity chose to write down, which is already a heavily filtered subset.
but think about everything humans knew but didn’t actually textualize like tacit knowledge, felt experience, etc. all of this is structurally absent from the training distribution.
which means the silence in ai outputs might be even more signal dense than in human ones, because the filter is double.
English