The bosses of OpenAI and Microsoft talk to The Economist

One reason the world’s corporate elite jet off to Davos each year is to check in on important relationships, be it with critical suppliers or big-spending customers. This year many are wondering about their relationships with Microsoft and OpenAI, the startup behind ChatGPT. The companies are the world’s most prominent purveyors of artificial intelligence (AI), which has the business world giddy. OpenAI exclusively licenses its technology to Microsoft. The software giant is busy injecting it into products from Word to Windows.

The relationship between the two companies is also under scrutiny. In November Sam Altman, OpenAI’s boss, was fired by his board, only to be reinstated days later. Satya Nadella, Microsoft’s chief, whose company reportedly owns 49% of the startup, supported Mr Altman during the ordeal. The kerfuffle still left many wondering about risks to what Mr Altman has called the “best bromance in tech”. When the pair sat down with The Economist in Davos on January 17th in their first joint public appearance since November, they were upbeat and, for the most part, singing from the same hymn sheet. Their partnership is “great” and “unbelievable”. They often remarked on how much they agree.

image: The Economist

One concurrence was that 2024 will be a big year for AI. Microsoft’s huge bet on the technology this month helped it to dethrone Apple as the world’s biggest firm (see chart). Today it is closing in on a value of $3trn. Its forthcoming quarterly earnings will give the first clear indication of how much corporate customers are willing to spend on AI. Although some observers have been underwhelmed by the progress made by OpenAI’s latest model, GPT-4, Mr Altman hints at new capabilities, such as greater ability to understand and produce audio. Mr Nadella says models will get better at all tasks, from writing essays to churning out computer code. “I really think the magic of this is the generality,” says Mr Altman.

The general-purpose nature of AI is one reason why Mr Altman thinks of the technology as “a new computer”. Mr Nadella sees it in similar terms. He argues that “since the PC, we have not had sort of the real driver of getting more things done with less drudgery.” Microsoft’s supply-chain team already use AI to help model the impact of their decisions, without having to wait for the finance department to do this at the end of the quarter.

AI’s ability to replace skilled workers, such as accountants, raises concerns about its impact on jobs. An IMF paper published on January 14th calculates that the technology could reshape labour markets. Those with college educations are both most exposed to disruption but also best positioned to reap the rewards of a new wave of innovation. Both Messrs Nadella and Altman are convinced that the technology will create more new jobs than it destroys. Mr Nadella thinks it may make the labour market more dynamic, by allowing people to learn new skills and switch jobs faster. That, he says, will cause some wages to go up and others to be “commoditised” (in other words, decline).

That disruption is likely to be all the more dramatic with the advent of artificial general intelligence (agi), which, if it is achieved, would be able to outperform humans on most intellectual tasks. AI doomers think this could generate all manner of ills, from economic chaos to a robot apocalypse. Nonetheless, producing AGI is the stated goal of OpenAI. Mr Altman describes progress towards this aim as “surprisingly continuous”. He likens it to the evolution of the iPhone, where no single new model represented a big leap but the technological advance from the first version to the latest one has been extraordinary. For that reason he expects the ado caused by the first AGI to be short-lived. “The world will have a two-week freakout and then people will go on with their lives,” he says.

Neither Mr Nadella nor Mr Altman will say when AGI might come around. Mr Nadella believes that by the time it does, its use will be regulated: “Nation states are absolutely going to have a say on…what is ready for deployment or not.” Mr Altman broadly agrees, but is a bit more circumspect. Regulators, he notes, will have to weigh the risks and capabilities of AI—as with aeroplanes, which create enormous benefits despite occasionally crashing. Likewise, AI’s “tremendous upside” means that halting progress would be a mistake. Safety is not a binary question of using or not using a technology; it is “the many little decisions along the way”. He points to the launch of GPT-4, which was pushed back by seven or eight months.

Mr Altman, ever the techno-optimist, insists that “technological prosperity is the most important ingredient to a much better future”. Mr Nadella, a corporate veteran, strikes a more businesslike note. He talks about the 20 meetings he had earlier in the day with executives from a range of industries, talking to them “about something that they are doing where I can have some input”. He is, in other words, firming up Microsoft’s relationships—as befits a big boss in Davos. 

Leave a Reply

Your email address will not be published. Required fields are marked *