StoryFirst seen

Microsoft builds trillion-parameter base models entirely in-house

State-of-the-art pretraining at the MAI lab gives Microsoft its own foundation-model pipeline and more control over its AI stack.

Archived Updated

Everything it touches

1 names 1 voices 1 source posts

The story’s neighborhood

NAMESVOICESPOSTS

Explore the graph, or choose a connection below.

Follow the evidence

The source posts

This brief comes from Tech Twitter’s edited headlines. The posts below are the sources grouped with the story; statements in them belong to their authors.

Mustafa SuleymanLead source
We now have state-of-the-art pre-trained base models in the 3T-parameter weight class, built entirely in-house at MAI. Post-training results coming soon -- stay tuned! Lots of details in our technical blog post.
Read original on X

How the headline evolved

  1. Microsoft builds trillion-parameter base models entirely in-house

  2. Microsoft builds trillion-parameter base models entirely in-house

  3. Microsoft builds trillion-parameter base models entirely in-house

  4. Microsoft builds trillion-parameter base models entirely in-house

  5. Microsoft builds trillion-parameter base models entirely in-house

  6. Microsoft trains trillion-parameter base models entirely in-house

  7. Microsoft trains trillion-parameter base models entirely in-house

  8. Microsoft trains trillion-parameter base models entirely in-house

  9. Microsoft trains frontier-class base models entirely in-house

  10. Microsoft trains frontier-class base models entirely in-house

  11. Microsoft's MAI lab trains frontier-class base models entirely in-house

  12. Microsoft's MAI lab trains frontier-class base models entirely in-house

  13. Microsoft's MAI lab trains frontier-class base models entirely in-house

  14. Microsoft's MAI lab trains frontier-class base models entirely in-house

  15. Microsoft's MAI trains state-of-the-art base models entirely in-house

Shared connections

Related stories