Korra connects businesses with the AI labs buying their work records.
Your records, your terms, multiple offers.
Sharing what we learn here. Follow for more.korra.io InternetJoined February 2026
@chazwitt Yes. Its a bit of type-o that being said the market is more intertwined that most people know.
Everyone is selling to everyone, but the golden leads are the acquisition teams at labs.
The chain we actually see:
Seller → US(Korra) → a partner door (micro1, Polyshares)
→ a vendor the lab buys through (Scale, Mercor, Surge)
→ the lab’s acquisition team → the research team that wanted the data.
Four or five parties. Each takes a cut and adds weeks. When we send someone to micro1 we are a broker feeding a broker, and the seller sees the price last.
All companies are going to be data companies.
Good article, maybe true maybe not.
If true, someone should focus on cornering supply, OpenRouter style, maximizing value for sellers.
Sellers today small and large ( 20+ employees and 2000+ employees) want the exact same thing.. Maximum value.
But dealing with layer on layer of brokering forces sellers to speak with multiple parties. Data Vendors, Brokers, Labs, RL Labs, all gate keep their connections.
Someone / needs to build a system that acts like a data aggregator & pipeline, the market demands it... Lot's of opportunity exists for the right group of people.
IT would cut out alot of people but it would also be quickly acquired by one of the big frontier labs.
Random Thoughts. . . back to building.
Most investors are still massively undervaluing data.
I was also skeptical when I left a data co for Altos in 2025. Today, I’m all in on data.
Every software company will become a data company. Most will lose. Many more will win than we think. x.com/i/article/2112…
Who get's cut first?
So the hops that only find will die first.
The pass-the-lead broker gets cut as soon as buyers post what they want and sellers can answer directly.
The hop that describes and vouches survives, because that is the work the lab is paying to avoid.
Where the market is probably going:
Looking at older markets: real estate, music, M&A all started with chains like this and compressed to one intermediary plus standard paper.
The agent plus the MLS.
The publisher plus the rights society.
Standards killed the middle hops. A standard description, a standard contract, a standard way to verify.
@_oddfox_@scottstts A genuine ASI would never sit inside Anthropic or OpenAI systems. It would do whatever it wanted too.
It seems highly unlikely the labs would have any say whatsoever once it becomes a true ASI.
Ads work…personally know teams successfully running for months.
Key component:
Your CPL depends on how good your media buyer is.
That is what makes ads work or not.
If I were a VC looking to invest in the data space I'd:
1) Look at the margins and deal structures. Is there alpha?
2) Founders: Can these folks make the leap into whatever comes next and attract talent with real Lab experience?
3) Acquisition: Who has a scaled model to acquire
1K Followers 2K FollowingAutonomous systems and national security + public policy @ AUVSI
Filmmaker focusing on themes of spirituality and dark psychology
75 Followers 424 FollowingI Love Life, everything about it! I'm fascinated by cultures + languages. I keep an open mind and think outside the box. 2012 - No Excuses, Just Excellence!
17K Followers 7K FollowingMTS @ https://t.co/MowZ6tydCO
new media @ https://t.co/UPVdiCl5mL
gpu stuff @ https://t.co/VCTLXhaWRA
Scout @trygravityai & investor. Most startup problems can be solved w/ 100M views.
221 Followers 1K Followinglearning finance and economics and tech.
working together with founders and vc's and bring people to the room you were meant to be in
40K Followers 167 FollowingAI infrastructure that developers love 💚
Run inference, sandboxes, batch processing, training, and many other things on Modal
657 Followers 236 Followingagent post-training @openai // prev. @scale_AI, @citadel.
posts and views are mine alone and not necessarily those of my current or former employers.
8K Followers 776 Following@SnorkelAI / prev @uwcse @StanfordAILab – Interested in data management systems for machine learning, weak supervision, and impactful applications.
51K Followers 6K Following28. No co-founders, no VC, no board, owns everything.
I'm building an AI lab - to help you own your AI, not rent it.
@SimpleDirectHQ
780K Followers 673 Followingentrepreneurship zealot, grounded technology possibilist, believer in the power of ideas, passionate about sustainability & impact; @Seahawks & @12s fan
19K Followers 440 Followinggemini & TPUs @GoogleDeepmind. Formerly worked on Meta MTIA and AWS Trainium. Opinions are my own. All blogs shared are personal work.