Skip to content

Research lab

AMI Labs

World model lab founded by Yann LeCun after leaving Meta; reported to have raised $1 billion.

Known aliases

  • AMI

Current stories

buildOne report1 publisher

PhAI Labs' JEPA-Anything outpredicts a standard JEPA in most tests across seven fields

PhAI Labs' JEPA-Anything splits a world model's prediction into several modules, cutting error 35% against a standard JEPA on a Pong task. Its reported comparisons use that one baseline, and its liver cancer pick has only lab and mouse data.

Publishers:the-decoder.com

Reality

Evidence40
Adoption
Not scored
Hype gap+30
Incentives
Not scored
Confidence35
investConfirmed2 publishers

Zenithon AI raises $10 million to cut fusion plasma simulations from months to seconds

London startup Zenithon AI has raised $10 million to train world models meant to replace slow physics simulations in fusion, rocketry and chip design. The company says its models check a million designs in the time one conventional simulation takes, with fusion energy developers among its early customers.

Reality

Evidence40
Adoption10
Hype gap+45
Incentives65
Confidence55
productOne report1 publisher

World model labs leave their own data supplier guessing at the spec

AMI Labs says it is still in a research and building phase and will not discuss product plans or a timeline. World Labs' Marble ships now as a way to build explorable scenes. A roadmap needs more than that.

Publishers:techcrunch.com

Reality

Evidence55
Adoption22
Hype gap+20
Incentives65
Confidence50
productOne report1 publisher

World Labs' Atlas turns one photo into a 3D scene a robot can be tested in

The capability claim is specific enough for a product team to test, with one image going in and up to a minute of 1440p video plus 3D splats coming out, though the camera-path win rests on a blind human study World Labs ran itself.

Publishers:siliconangle.com

Reality

Evidence32
Adoption12
Hype gap+34
Incentives78
Confidence40
investOne report1 publisher

Three rounds, $2.31B, and one experiment where blocking 1% of Manhattan broke the model

Techfundingnews frames the world model race as a $16B bet. The named rounds account for $2.31B of it, and the case against the architecture rests on a generalisation failure capital cannot fix.

Reality

Evidence28
Adoption16
Hype gap+46
Incentives86
Confidence34