Microsoft this week gained an exclusive license to OpenAI’s GPT-3, the state-of-the-art language model garnering attention across the tech industry. Other companies will still be able to access the model through an Azure-hosted API, but only Microsoft will have access to GPT-3’s code and underlying advances. The deal follows Microsoft’s $1 billion investment last year in San Francisco-based OpenAI, which consists of the OpenAI Inc nonprofit founded four years ago and the for-profit OpenAI LP.
The implications of giving a tech giant such as Microsoft an exclusive license to GPT-3 raises questions and potential concerns. MIT Technology Review said this week that OpenAI was “supposed to benefit humanity,” and now “it’s simply benefiting one of the richest companies in the world.”
GPT-3 is the largest language model in the world and can quickly create human-like written text using machine learning technologies. Earlier this month The Guardian published an essay written from scratch by GPT-3. “Overall, it took less time to edit than many human op-eds,” said the site’s human editors.
With its exclusive license, Microsoft CTO Kevin Scott said the company aims to “expand our Azure-powered AI platform in a way that democratizes AI technology, enables new products, services and experiences, and increases the positive impact of AI at Scale.”
“Our mission at Microsoft is to empower every person and every organization on the planet to achieve more, so we want to make sure that this AI platform is available to everyone – researchers, entrepreneurs, hobbyists, businesses – to empower their ambitions to create something new and interesting,” Scott wrote in a blog post.
GeekWire reached out to AI experts and other tech insiders in Seattle to get their take on Microsoft’s exclusive license.
“OpenAI should be renamed ClosedAI — for all intents and purposes they are a for-profit company,” said Oren Etzioni, CEO of the Allen Institute for Artificial Intelligence (AI2).
Pedro Domingos, a University of Washington computer science professor and author of The Master Algorithm, said the deal makes sense logistically since Microsoft is better set up for licensing that OpenAI.
“OpenAI is still working to benefit humanity; it has had to become more commercial in order to be able to compete with the corporate labs, but I don’t begrudge them that,” he said.
Domingos added that he doesn’t think GPT-3 is the next big thing.
“It’s a nice demonstration of what large-scale machine learning can do, and I’m curious to see what applications of it people come up with, but it’s not a fundamental advance,” he said. “It’s just scaling up pre-existing technology, as OpenAI’s paper on it makes clear.”
Others are more bullish. Greg Gottesman, managing director at Seattle startup studio Pioneer Square Labs, said he and his colleagues are building multiple prototypes using GPT-3.
“The results are more often shockingly good than entertainingly bad,” he said. “Based on these early prototypes, we believe GPT-3 is a leap forward in enabling AI applications, and our hope is that the platform will remain open.”
Etzioni said his lab is thinking about how GPT-3 can extend to the visual realm; how the models can provide accurate information, not just compelling text; and how to fight misinformation that could be catalyzed by AI advances such as GPT-3.
“GPT-3 has remarkable capabilities and will lead to numerous applications and an even more vigorous mine-is-bigger-than-yours model arms race,” he said. “I can’t wait to see how Google and Amazon respond, and don’t forget China.”
Microsoft followed its $1 billion investment by building one of the world’s top supercomputers for the exclusive use of OpenAI.
OpenAI was formed in 2016 by Elon Musk, the Tesla and SpaceX CEO; Sam Altman, the former Y Combinator president; Ilya Sutskever, OpenAI’s chief scientist; and Greg Brockman, the former Stripe CTO. Musk, who has sounded the alarm over the risks of AI, said last year that he was no longer involved in OpenAI.