Close Menu
Tech News VisionTech News Vision
  • Home
  • What’s On
  • Mobile
  • Computers
  • Gadgets
  • Apps
  • Gaming
  • How To
  • More
    • Web Stories
    • Global
    • Press Release

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

Trending Now

Amnesty International raises exclusion concerns about government’s ‘unchecked’ use of AI

10 July 2025

Stapler Behind GameStop’s Infamous Switch 2 Launch Disaster Up for Auction — and the Bidding Is Now Over $100,000

10 July 2025

Day three Prime Day 2025 deals are just as good as day one

10 July 2025
Facebook X (Twitter) Instagram
  • Privacy
  • Terms
  • Advertise
  • Contact
Facebook X (Twitter) Instagram Pinterest VKontakte
Tech News VisionTech News Vision
  • Home
  • What’s On
  • Mobile
  • Computers
  • Gadgets
  • Apps
  • Gaming
  • How To
  • More
    • Web Stories
    • Global
    • Press Release
Tech News VisionTech News Vision
Home » A New Kind of AI Model Lets Data Owners Take Control
What's On

A New Kind of AI Model Lets Data Owners Take Control

News RoomBy News Room9 July 2025Updated:9 July 2025No Comments
Facebook Twitter Pinterest LinkedIn Tumblr Email

A new kind of large language model, developed by researchers at the Allen Institute for AI (Ai2), makes it possible to control how training data is used even after a model has been built.

The new model, called FlexOlmo, could challenge the current industry paradigm of big artificial intelligence companies slurping up data from the web, books, and other sources—often with little regard for ownership—and then owning the resulting models entirely. Once data is baked into an AI model today, extracting it from that model is a bit like trying to recover the eggs from a finished cake.

“Conventionally, your data is either in or out,” says Ali Farhadi, CEO of Ai2, based in Seattle, Washington. “Once I train on that data, you lose control. And you have no way out, unless you force me to go through another multi-million-dollar round of training.”

Ai2’s avant-garde approach divides up training so that data owners can exert control. Those who want to contribute data to a FlexOlmo model can do so by first copying a publicly shared model known as the “anchor.” They then train a second model using their own data, combine the result with the anchor model, and contribute the result back to whoever is building the third and final model.

Contributing in this way means that the data itself never has to be handed over. And because of how the data owner’s model is merged with the final one, it is possible to extract the data later on. A magazine publisher might, for instance, contribute text from its archive of articles to a model but later remove the sub-model trained on that data if there is a legal dispute or if the company objects to how a model is being used.

“The training is completely asynchronous,” says Sewon Min, a research scientist at Ai2 who led the technical work. “Data owners do not have to coordinate, and the training can be done completely independently.”

The FlexOlmo model architecture is what’s known as a “mixture of experts,” a popular design that is normally used to simultaneously combine several sub-models into a bigger, more capable one. A key innovation from Ai2 is a way of merging sub-models that were trained independently. This is achieved using a new scheme for representing the values in a model so that its abilities can be merged with others when the final combined model is run.

To test the approach, the FlexOlmo researchers created a dataset they call Flexmix from proprietary sources including books and websites. They used the FlexOlmo design to build a model with 37 billion parameters, about a tenth of the size of the largest open source model from Meta. They then compared their model to several others. They found that it outperformed any individual model on all tasks and also scored 10 percent better at common benchmarks than two other approaches for merging independently trained models.

The result is a way to have your cake—and get your eggs back, too. “You could just opt out of the system without any major damage and inference time,” Farhadi says. “It’s a whole new way of thinking about how to train these models.”

Share. Facebook Twitter Pinterest LinkedIn Tumblr Email

Related Posts

OpenAI reportedly preparing AI-powered web browser to challenge Google Chrome

10 July 2025

X’s CEO is out after failing at basically everything she claimed she wanted

10 July 2025

Hyundai Reveals the ‘Lightspeed’ Ioniq 6 N

10 July 2025

Amnesty International raises exclusion concerns about government’s ‘unchecked’ use of AI

10 July 2025
Editors Picks

Nintendo Ditching Switch Game Vouchers, Ending a Good Way to Save Money on Switch 1 Games and Switch 2 Upgrades

10 July 2025

X’s CEO is out after failing at basically everything she claimed she wanted

10 July 2025

Hyundai Reveals the ‘Lightspeed’ Ioniq 6 N

10 July 2025

Everything You Need to Know Before Buying a Fitness Wearable

10 July 2025

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

Trending Now
Tech News Vision
Facebook X (Twitter) Instagram Pinterest Vimeo YouTube
  • Privacy Policy
  • Terms of use
  • Advertise
  • Contact
© 2025 Tech News Vision. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.