Igor Zalutski
@IgorZIJ
Building homes for agents
prev @PalantirTech @fitbit
2.8K Following 2.3K Followers
Yes what could possibly have happened
Linux currently has an all-time-high desktop market share of 7.53% as of July 2026. But what exactly happened over the last two months? In May 2026, Linux’s market share was 3.81%, so how did Linux gain approximately 3.72 percentage points in just two months?
Show more
Reminder that the best place to be in SF tomorrow evening is @UseCorgi’s new cafe in dogpatch for the @opencomputerHQ agent builder happy hour.
Excited for this!
last night was super fun despite me being horribly ill
@p0 really knocked it out the park with getting us 🧋 it felt so SO good to get the London AI community back together after a summer hiatus.
so many cool things happened last night, you had to be there.
Show more
pleasant surprise this morning to see a bunch of traffic on the @opencomputerHQ website coming from @bentossell's newsletter!
LFG!
tbh. I wish more people didn't lean on the agentic loop 100%; instead, they should lean on @temporalio esq workflow engines as stages. There's an art in deciding whether to utilise the LLM loop vs. a deterministic workflow. You can do both; the magic is in combining and applying both ingredients in the right manner.
Show more
Demoing OpenComputer Serverless Agents at the AI Demo Nights in London tonight by @Andy_AJT - come say hi!
Introducing
Deploy your agent as a function. Get a computer for it.
- a real Linux machine per session
- ffmpeg, chromium, git, install anything
- durable: hibernate + resume
- your keys never enter the runtime
The cloud of the last decade was built for apps. The next one is for trillions of always on agents. @opencomputerhq is their home.
Show more
Every day I'm reminded that I spent my youth in meetings with a stop watch arguing about how to shave fractions of a second off execution of programs.
Then came dialup internet.
Today the most used program is so slow they rebranded slow as "agentic" and everyone is thrilled!
Show more
Tomorrow we ship the thing we've been hinting at for MONTHS.
I won't say much yet, except that it's the biggest launch we've ever done, beta users have been loving it, and I have not slept properly in two weeks 😅
Friday we're throwing the Agent Builder Happy Hour at @UseCorgi's new Dogpatch cafe.
Pull up.
Luma to RSVP:
Show more
i really don't want to use your agent, i want to use my agent to use your thing.
No loyalty is a good thing.
It means the space isn’t commoditised, and there’s room for innovation & that users are open to switching between tools.
Loyalty is in hyper commoditised spaces typically - airlines, coffee shops, restaurants etc.
Show more
More BTS of @motatoeshq KILLING IT
Something something “home for agents”
BTS of the @opencomputerHQ launch video.
Big things coming!
Why agents should be defined as functions & not json or YAML
Heroku nailed prosumer adoption (before tiktok, ugc etc) but unfortunately never cracked enterprise. The hobbyists loved it and stayed hobbyists.
I played with the @ElevenLabs' managed agent platform (For voice agents), I think that ceiling is gone.
The loop is the kinda the same
> practioner tinkers
> "holy shit"
> tells their friends.
But now the friend teller is a staff eng at Stripe/Coinbase/Doordash/Airbnb, and the tool follows them into prod.
Prosumer is the enterprise wedge now.
Show more
The man, the myth, the legend @vmg does it again:
read: that compute is needed more for Astra inference
=> Astra is likely very good
we'll know in 2 days i guess
We have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us. Model progress is now extremely rapid, and we always said we would take action if we felt that model capabilities were outstripping the pace of safety and alignment.
We care very deeply about AI safety. We believe the entire field will have to coordinate on shared safety standards, but will act unilaterally in the meantime.
We expect confidence in safety to increasingly set the pace of AI progress. We are optimistic about the alignment work we are doing, and we remain committed to making frontier capabilities widely available.
Show more
