Jannah Theme License is not validated, Go to the theme options page to validate the license, You need a single license for each domain name.
Politics

With the return of Sam Altman, AI moves from idealism to pragmatism

Nature

Listen to this story.
Enjoy more audio and podcasts on iOS Or Android.

Your browser does not support the element

ESOLD BY TECH’S rapidly evolving standards, last week in the world of artificial intelligence (AI) made your head spin. On November 17, the Open Board of DirectorsAI expelled Sam Altman, the CatGoogle Tag-the manufacturer’s boss. On November 20, Mr. Altman was offered refuge at Microsoft, the startup’s biggest backer. Same day almost the entire OpenAIThe company’s 770 employees signed a letter threatening to resign unless the board members who fired Mr. Altman reinstated him and resigned. On November 21, Mr. Altman returned to his old job. The heads have therefore returned more or less to their starting point. Or did they?

In fact, the OpenAI saga marks the beginning of a new, more adult phase for the AI industry. For OpenAI, the triumphant return of Mr. Altman could strengthen his ambitions. For Microsoft, which supported Mr. Altman in his difficult times, this episode could lead to greater influence on AIthe hottest startup. For AI around the world, it could herald a wider abandonment of academic idealism in favor of greater commercial pragmatism. And for users of this technology, it could hopefully lead to more competition and more choice.

To understand all of these implications, start with what happened. OpenAIThe board fired Mr. Altman for not being “consistently forthright in his communications.” One factor that may have influenced the decision was disagreement over whether the OpenAI had found the right balance between speed and safety of its products. Insiders say OpenAI had made a breakthrough that allowed models to better solve problems without additional data. This scared co-founder and board member Ilya Sutskever. Helen Toner, a board member affiliated with Georgetown University, had published an academic article exposing what she saw as flaws in OpenAIthe approach of AI security. On November 21, the New York Times reported that Mr. Altman, worried about negative press, had decided to oust Ms. Toner. Concerns have also been raised about Mr. Altman’s side projects, including a project AI-a semiconductor company that sent him to the Persian Gulf to court billions in Saudi money.

Ultimately, it was Ms. Toner and three other board members who ousted him. The sixth director, Greg Brockman, was also stripped of his board seat, then resigned in solidarity with Mr. Altman. Both found support at Microsoft, which announced it would create a new internal service AI laboratory that they would direct. Microsoft has also committed to hiring the rest of OpenAI‘steam. We may never know if this project was serious or not. But it gave Mr. Altman enormous bargaining power in negotiations over his return to the Open.AI. On November 20, while these negotiations were underway, Satya Nadella, the tech giant’s chief executive, said, “No matter where Sam is, he’s working with Microsoft.”

The deal reached by Mr. Altman and those who ousted him will transform the OpenAI, starting with the table. Ms. Toner and Mr. Sutskever are absent. The same goes for Tasha McCauley, a tech entrepreneur. All three supported Mr. Altman’s firing. Mr. Brockman and, at this time, Mr. Altman will not be returning. Of the six from before the chaos, only Adam D’Angelo, the founder of Quora, a question and answer site, remains. He will be joined by heavyweights, starting with Bret Taylor, former co-CEO of Salesforce, another major software company, and Larry Summers of Harvard University, who served as Bill Clinton’s Treasury secretary. THE Edge, an online publication, reported that the new board would aim to expand to nine members; Microsoft should get a seat and Mr. Altman could get it back.

New directors should make OpenAI, which is structured as a for-profit entity within a more business-oriented, nonprofit entity. Mr. Taylor and Mr. Summers are highly regarded individuals with extensive board experience. Their opinions on AI safety is not known. But they might be more receptive than Ms. Toner and Ms. McCauley to Mr. Altman’s empire-building ambitions. The same already seems to be true for OpenAIthe workforce. One employee reports that the startup’s staff, who “came together” during the upheaval, will become even more loyal to Mr. Altman and, perhaps, more willing to pursue his business vision. Work on the company’s most powerful model to date, Google Tag-5, which seemed to have slowed down in recent months, should now move forward at full speed.

The sour taste left by the imbroglio may nevertheless persist. It was not, in the words of one eminent AI investor, a “confidence event”. That’s an understatement. The morning of November 17 OpenAI was close to closing a tender offer led by Thrive Capital, a venture capital firm, that would value the startup at $86 billion. The offer has been suspended. Although it appears to be back, investors in the secondary market for startup stocks remain cautious. Worse still, if Mr. Altman and Mr. Sutskever do not reconcile, OpenAI could lose one of the most respected men in the world AI the spirits.

Microsoft’s fortunes seem more secure. While openAIMicrosoft’s brand took a hit, not Microsoft’s. The software giant probably prefers to have OpenAI at arm’s length rather than Mr. Altman and his boffins close to his chest. In terms of temperament, Mr. Altman and Mr. Brockman are not natural fits for one of the world’s largest companies; many observers doubted that either would have stayed at Microsoft for long.

Recreate openAI In-house production would also have slowed progress in the technology in the short term, says Mark Moerdler of Bernstein, a broker. Many openAI The employees said privately that they would prefer to join another company than Microsoft, even though they signed the petition threatening to follow Mr. Altman there. Mr. Nadella does not seem terribly disappointed with the result. Microsoft’s stock price, which fell 2% on news of Mr. Altman’s dismissal, recouped all of those losses. On November 22, its market value reached an all-time high of $2.8 trillion.

image: The Economist

What about the rest of the AI industry? OpenAI is the undisputed leader of the AI race (see table). A survey by Retool, a startup, found that 80% of software developers said they use OpenAImore often than those of competing modelers. CatGoogle Taga chatty app whose launch a year ago has become openAI become a household name, receives 60% of web traffic to the top 50 websites for this type of “generation” AI. In October, the company had annualized revenue of $1.3 billion.

Even if openAI moves faster under new management, it will face more competition. A AIThe venture capital-focused venture capitalist compares this moment to the implosion earlier this year of Silicon Valley Bank, which taught many startups not to put all their eggs in one basket. As the Altman drama unfolded, more than 100 OpenAI customers contacted Anthropic, a rival model maker, according to the Information, an online publication. Some have tapped Cohere, another startup, and Google’s cloud unit, which has invested in Anthropic. The cloud arm of Amazon, another Anthropic backer, has set up a team to work with the mixers.

Events at OpenAI are a dramatic manifestation of a broader division within Silicon Valley. On one side are the “condemned”, who believe that if nothing is done, AI poses an existential risk to humanity and therefore calls for stricter regulations. Opposite them are the “baby boomers”, who downplay fears of a have apocalypse and highlight its potential to boost progress. This split partly reflects philosophical differences. Many in the doomsday camp are influenced by “effective altruism”, a movement that worries that Have could wipe out humanity. Baby boomers embrace a worldview called “efficient accelerationism,” which opposes the development of AI should be accelerated.

Mr. Altman appeared to have sympathy for both groups, publicly calling for “safeguards” to prevent have safe by pushing Openhave develop more powerful models and launch new tools, such as an app store allowing users to create their own chatbots. Today it looks much more boomerang, just like the majority of OpenAIthe workers who wanted him back. The pessimists are in retreat.

This will worry policymakers, who are struggling to show they are taking the risks seriously. In July, President Joe Biden’s administration gave a boost to seven major modelers, including Google, Meta, Microsoft and Open.haveto make “voluntary commitments” so that their have products inspected by experts before making them public. On November 1, the British government asked a similar group to sign another non-binding agreement allowing regulators to test their haves for reliability and harmful capabilities, such as endangering national security.

A few days earlier, Mr. Biden had issued a more biting executive order. This obliges anyone have companies build models above a certain size – defined by the computing power required – to inform the government and share the results of its security tests. As baby boomers take over in Silicon Valley, the White House’s model inspectors should expect to have their hands full.

Read the rest of our articles on artificial intelligence

To stay on top of the biggest business and technology news, sign up for Bottom Line, our weekly subscribers-only newsletter.

Nature
politics New Gb1

Not all news on the site expresses the point of view of the site, but we transmit this news automatically and translate it through programmatic technology on the site and not from a human editor.
Back to top button