The $10 billion investment that Microsoft is considering in San Francisco-based research firm OpenAI looks set to become the defining deal for a new era of artificial intelligence.
If the US software giant is right about the technology’s far-reaching implications, it could also lead to a realignment in the AI world as other tech groups race for their place in the new field of generative AI.
OpenAI made global headlines last month with the launch chatan artificial intelligence system that can respond to queries and produce text in language that looks natural.
But Microsoft executives believe the technology behind the service will soon have a deeper impact throughout the tech world.
“these [AI] “Models will change the way people interact with computers,” said Eric Boyd, Head of Artificial Intelligence Platforms at Microsoft. He added that speaking to a computer as naturally as a person would revolutionize the daily experience of using technology.
“They understand your intent in a way that wasn’t possible before and they can translate that into computer procedures,” Boyd said in an interview with the Financial Times before news of the potential deal.
Microsoft’s potential investment, first reported in last week’s Semaphore newsletter and confirmed by two people familiar with the situation, would take a significant minority stake that would value OpenAI, after the investment, at $29 billion. Microsoft declined to comment.
Potential historical investment comes as venture capitalists Rushing to support the latest trend of artificial intelligence At a time when past investment fads like blockchain and cryptocurrency have faded away.
Microsoft She made her first billion dollar investment at OpenAI in 2019, resulting in its role as the technology platform for the company’s high-demand AI models and giving it first rights to commercialize its technology.
The software giant has already used OpenAI technology in a number of its own products, though its executives say this only scratches the surface of what comes next.
Cloud customers have been able to pay for access to GPT-3, an AI model for text generation, since 2021. Dall-E 2 – part of wave of image generation systems What revolutionized the world of artificial intelligence last year – is the basis of a recent graphic design product from Microsoft called Designer, and also made available through the Bing search engine.
Meanwhile, Codex, a system that prompts software developers for suggestions about which lines of code to write next, has been made into a product by GitHub, Microsoft’s service for developers.
According to AI experts, the speed with which AI tools move from advanced search to everyday products may be unparalleled in the history of technology. Codex was only introduced in an OpenAI paper in mid-2021, but within a year Microsoft had turned it into a commercial subscription service.
According to GitHub CEO Thomas Dohmke, 40 percent of the code created by developers who use the service, called Copilot, is automatically generated by an AI system, cutting the time it takes to generate new code in half — a huge jump in efficiency after a decade. A largely ineffective effort to boost developer productivity.
“It’s an amazing productivity statistic,” Dumke said.
Much of OpenAI’s technology stems from creating so-called large language models, which are trained on huge amounts of text. Unlike previous forms of machine learning — which have dominated AI over the past decade — this technology has led to systems that can be used in a wide variety of circumstances, enhancing their commercial value.
“The real strength of these models is that they have the ability to do many different tasks at the same time,” said Microsoft’s Boyd. He added that it makes it possible to do what is called “no shot” learning – using AI to perform new tasks without having to train them.
Google and other tech giants, as well as a number of startups, have invested resources in creating giant AI models like this one. But ever since GPT3 stunned the AI world in 2020 with its ability to produce large batches of text on demand, OpenAI has set the pace for a series of eye-catching public demonstrations.
Microsoft executives are looking to use the technology in a wide range of products. Speaking at a company event late last year, CEO Satya Nadella predicted that generative AI would lead to “a world where everyone, regardless of their profession, can ‘get powered by technology’ for everything they do.”
Generative AI is set to become a core part of “productivity” applications like Microsoft Office, said Oren Etzioni, an advisor and board member at A12, the AI research institute set up by Microsoft co-founder Paul Allen.
He said all workers will eventually use productivity software that presents them with relevant information, scans their work and offers to automatically generate content.
The potential disruptions this could cause in the software world are not lost on Microsoft’s competitors, who see the technology as a rare opportunity to break into markets dominated by Big Tech.
Emad Mostaq, head of London-based Stability AI, which hit a huge success last year with the launch of the open source image creation system, claimed that his organization was building a “PowerPoint killer” – an AI tool designed to make it easier to create presentations than a widely used Microsoft app. .
This makes the move both defensive and offensive for Microsoft, as it tries to protect established products like Office while facing a stronger challenge in markets like Internet search.
Meanwhile, with its potential investment in OpenAI, Microsoft is also trying to use its technological and financial strength to establish itself as the main platform on which the next era of artificial intelligence will be built.
The power of cloud computing [OpenAI] One of the company’s investors said that the needs are beyond the ability of a startup or a venture capital investor to support. This person added that this meant that OpenAI had no choice but to seek financial support from one of the few tech giants.
Microsoft has sought to use its first investment in OpenAI to get a head start, building what it describes as a supercomputer to train giant AI models for the research company. The same technology platform is now also used by Meta, the parent company of Facebook, for its work on artificial intelligence.
Nadella recently claimed that the head start he got from working with OpenAI means that the calculations performed by the AI supercomputer cost about half as much as its biggest competitors. Any cost advantage could be primary: Analysts at Morgan Stanley estimate that the higher cost of natural language processing means that answering a query using ChatGPT costs about seven times as much as a normal Internet search.
Microsoft’s biggest cloud computing competitor is also seeking to align itself with some of the most promising generative AI companies, though none of the other startups in the field are producing AI models of the scale or scope of OpenAI.
Amazon’s cloud division has a three-year deal to serve as the computing platform for Stability AI. CohereAI, another AI startup founded by three Google researchers, reached an agreement in 2021 to use the research company’s computing platform to train their AI.
If a few tech giants become the central platforms — and investors in them — for startups building the next generation of AI technology, that could spark concern among regulators.
A person familiar with Microsoft’s investment plans acknowledged that its alliance with OpenAI would likely come under close scrutiny, but added that the minority investment should not spark any regulatory interference.
But as owners of cloud computing platforms need to support the coming era of generative AI, it seems inevitable that big tech will have an important say in what comes next.