Microsoft hires former Ai2 executive Ali Farhadi and key researchers for Suleyman’s AI team

Microsoft is hiring a group of top AI researchers from the Seattle-based Allen Institute for AI and the University of Washington, including former Ai2 CEO Ali Farhadi, GeekWire has learned.

Farhadi, Hanna Hajishirzi, and Ranjay Krishna are expected to join Mustafa Suleyman’s organization at Microsoft while maintaining their faculty positions at the UW’s Allen School of Computer Science and Engineering. Also joining is Sophie Lebrecht, former CEO of Ai2.

The move follows Farhadi’s departure from Ai2, which was announced on March 12. Farhadi had led the Seattle-based research institute for more than two and a half years.

Suleyman, the CEO of Microsoft AI, narrowed his focus last week from overseeing Copilot’s consumer-focused products to leading the Microsoft Superintelligence team.

The hiring comes as Microsoft works to reduce its reliance on OpenAI for frontier AI models, competing with Amazon, Google, and others. Suleyman’s Superintelligence Team, formed in November, is part of a broader campaign to develop advanced base models.

Microsoft has already hired researchers from Google DeepMind, Meta, OpenAI, and Anthropic, and the addition of the Ai2 team with UW will bring deep expertise in open model development and training efficiency – where Ai2 punches above its weight.

Supported by NSF and Nvidia

The exit marks a significant loss for the Ai2 group, which was founded in 2014 by the late Microsoft co-founder Paul Allen. Hajishirzi is co-lead of the OLMo open source language modeling project and principal investigator of the $152 million, five-year effort supported by the National Science Foundation and Nvidia to build open AI models for scientific research.

He represented Ai2 several times last week at Nvidia’s GTC conference in San Jose, including a panel on the future of open source models alongside Nvidia CEO Jensen Huang.

Krishna led the development of Ai2’s Molmo multimodal models, among other projects. He also presented at Nvidia’s conference last week on behalf of the institute.

Farhadi, a computer vision specialist, founded the Ai2 spinout Xnor.ai, which was acquired by Apple in 2020 for approximately $200 million. He led machine learning efforts at Apple before returning to lead Ai2 as CEO in July 2023.

Ai2’s interim CEO, Peter Clark, acknowledged the move in a statement, saying the agency remains committed to its mission and collaboration with NSF and Nvidia, including the OMAI program.

“These programs are supported by a broad, experienced team with the knowledge and continuity needed to move this project forward,” Clark said. “We are confident in our ability to build on the strong foundation already in place and to increase the impact of these efforts in the coming months.”

He added that the institute is “grateful for the leadership and contributions of Ali, Hanna, Ranjay, and others” in advancing Ai2’s work, and wishes them the best.

In a post about the hire on LinkedIn, Suleyman praised Farhadi for leading Ai2 to release more than 100 models in one year and called Hajishirzi “one of the most cited natural language processing researchers in the world, full stop.”

Suleyman described Lebrecht as balancing Ai2’s work and open source efforts, noting that he also founded the AI ​​company Neon Labs and holds a PhD in cognitive neuroscience from Brown University.

He said they will help pursue Microsoft’s mission of “human intelligence: safer, more manageable, more efficient AI systems for serving humanity and our most pressing problems.”

When news broke earlier this month that Farhadi was leaving, Ai2 board chairman Bill Hilf told GeekWire that Farhadi wanted to do research at the extreme frontier of AI, where for-profit companies are spending billions on training more advanced models.

At the time, Hilf said the board should have evaluated whether the nonprofit’s dollars were better spent trying to keep pace, acknowledging that competing with tech giants on the largest scale of model development has been incredibly difficult.

Changes to Ai2 funding facts

Behind the scenes, changes in Ai2’s funding situation also played a role in the exit, according to people with knowledge of the situation.

Ai2 was originally sponsored by Allen’s Vulcan Inc. and was later funded by his legacy. Its main supporter is now the Science and Technology Fund, a $3.1 billion foundation created under Allen’s instructions and launched publicly in August, focused on applying science and technology to problems in areas that align with Allen’s passions, including AI, bioscience, and the environment.

FFST, led by CEO Dr. Lynda Stuart, a physician and scientist who once led the Institute for Protein Design at the UW, favors the use of AI in addition to the expensive work of frontier models.

In addition, while all Ai2 programs by 2026 are fully funded, these people said, FFST is moving from giving Ai2 all the money annually to a proposal-based program, with future support expected to favor real-world applications of AI in addition to building open-source base models. The shift helps explain the flow of researchers focused on model development.

A spokesperson for the Science and Technology Fund said “the mission and purpose of Ai2 remains the same” and that strategies for the wider FFST program are still being developed.

Farhadi, Hajishirzi, and Krishna are researchers who focus their work on building and developing AI models. The Microsoft Superintelligence team, backed by billions in computing investment, provides the resources and authority to pursue that mission on a much larger scale.