It’s easy to worry that machines have taken over. Companies such as IBM and British telecoms company BT have cited artificial intelligence as a reason for layoffs. And new tools like ChatGPT and DALL-E help everyone understand what’s wrong. capabilities of artificial intelligence itself. A recent study by OpenAI (the startup behind ChatGPT) and University of Pennsylvania researchers found that for about 80% of jobs, the technology behind such tools automates at least 10% of the tasks. I conclude that I can.
“Anyone I talk to, super-smart people, doctors, lawyers, CEOs, other economists, their brains are like, ‘Oh, how can we replace what humans are doing with generative AI?’ ?’” said Professor Erik Brynjolfsson of the university. Stanford Human-Centric AI Lab.
But it’s not the only option, he said. “Another thing I’d like people to do more of is think about what new things can be done now that haven’t been done before. Obviously, that’s a much harder question.” is also “the most valuable place,” he added.
How technology makers design AI tools, how business leaders use them, and how policymakers regulate AI tools will determine how generative AI will ultimately impact jobs. says Brynjolfsson and other economists. And not all options are necessarily bleak for workers.
AI can complement human labor rather than replace it. For example, many companies are leveraging AI to automate their call centers. But Fortune 500 companies that provide business software are instead using tools like ChatGPT to give employees real-time suggestions on how to engage with customers. Brynjolfsson and his co-authors compared call center employees who used the tool with those who did not. They found that the tool increased productivity by an average of 14%, with most of the gain coming from low-skilled workers. The group that used this tool also had higher customer sentiment and lower employee turnover.
Massachusetts Institute of Technology economics professor David Autor said AI could potentially be used to provide “immediate expertise” in jobs such as healthcare delivery, software development, law and skilled repair. Stated. “This provides an opportunity to enable more workers to do valuable jobs that rely on some of their expertise,” he said.
Employees can focus on different tasks. According to analysis by James Bessen, a researcher at Boston University Law School, the number of bank tellers has increased as ATMs have automated the task of withdrawing cash and taking deposits. This was partly because banks opened more branches, while fewer workers were needed for bank branches, while the costs of opening branches became cheaper. But banks have also changed job descriptions. Since the ATM, tellers have focused less on counting cash and more on building relationships with customers selling products such as credit cards. Few jobs can be fully automated with generative AI. However, using AI tools for some tasks could augment workers’ work on tasks that cannot be automated.
New technology can lead to new jobs. Agriculture employed nearly 42 percent of the labor force in 1900, but had fallen to just 2 percent in 2000 due to advances in automation and technology. The sharp decline in agricultural employment did not lead to an increase in unemployment. Instead, technology has created many new jobs. Early 20th-century farmers would never have imagined computer coding, genetic engineering, and trucking. In an analysis using census data, Autor and his co-authors found that 60 percent of today’s occupational specialties did not exist 80 years ago.
Of course, there is no guarantee that a worker will be eligible for a new job or that they will get a good job. None of this happens out of the blue, says Daron Acemoglu, an economics professor at the Massachusetts Institute of Technology and co-author of Power and Progress: Our 1,000-Year Struggle Over Technology & Prosperity.
“If we make the right choices, we can create new types of jobs, which are crucial both for wage growth and for real productivity gains,” Acemoglu said. . “But nothing like this is even possible if we don’t make the right choices.” — Sarah Kessler
in case you missed
Martha’s exemplary behavior. Lifestyle entrepreneur Martha Stewart became the oldest person to appear on the cover of Sports Illustrated’s swimsuit issue this week. Stewart, 81, told The Times that posing with her confidence was a “huge challenge,” but two months of Pilates helped her. She is not the first person over the age of 60 to receive this honor. Last year, Elon Musk’s mother May Musk made the cover at age 74.
TikTok block. Montana became the first state to ban Chinese short-video apps, banning app stores from offering TikTok within state borders from January 1. Enforcing the ban is expected to be difficult, and TikTok users in the state are suing the government. The move violates First Amendment rights and provides a glimpse of the potential backlash if the federal government tries to block her TikTok nationwide.
Banker responsibility game. Former Silicon Valley Bank CEO Greg Becker, in his first public comment since the bank’s collapse in March, blamed “rumours and misunderstandings” for the run. Mr. Becker and former executives at the failed Signature bank told a Senate committee investigating their role in the failure of the bank that they would not return their million-dollar paychecks.
A Brief History of Constraint-Seeking Tech CEOs
When OpenAI chief executive Sam Altman testified before Congress this week calling for regulation of generative artificial intelligence, some lawmakers hailed it as a “historic” move. In fact, calling on lawmakers for new rules is a move straight out of tech industry strategy. Silicon Valley’s most powerful executives traveled to Washington at length to demonstrate their commitment to the rules, to shape them while also tirelessly unleashing some of the world’s most powerful and innovative technologies. It’s here.
One reason, Bruce Melman, a political consultant and former Bush administration technology policy officer, told DealBook that federal regulations are much easier to manage than state-by-state regulations. about it. Clarifying regulations also increases investor confidence in the sector, he added.
This strategy seems sensible, but if history is a helpful guide, reality can be more troublesome than rhetoric.
-
In December 2021, bankrupt crypto exchange FTX founder Sam Bankman-Fried was one of six executives who testified before the House about digital assets and called for regulatory clarity. . He told lawmakers that his company had just submitted a proposal for a “unified joint system.” A year later, Bankman his Freed business went bankrupt and he faced charges of criminal fraud and illegal campaign contributions.
-
In 2019, Facebook founder Mark Zuckerberg wrote an op-ed in The Washington Post titled “The Internet Needs New Rules,” given the company’s failures in content management, election integrity, privacy and data stewardship. contributed. Two years later, independent researchers found that misinformation is more prevalent on the platform than it was in 2016, despite the company’s billions of dollars spent eradicating it.
-
In 2018, Apple Chief Executive Tim Cook said he generally opposed regulation but supported stricter data privacy rules, adding, “It’s time for a lot of people to think about what they can do.” said. But in order to keep its business in China, one of its largest markets, Apple handed over most of its control over customer data to the government as part of its operating requirements in China.
Buzzword of the week: “Argospeak”
Platforms like TikTok, Facebook, Instagram, and Twitter use algorithms to identify and moderate problematic content. A language code was developed to circumvent these digital moderators and allow free exchange on taboo topics. It is called “Argospeak”.
“Online, the language arms race is heating up, but it’s not clear who’s winning.” Roger J. Kreutz, professor of psychology at the University of Memphis, writes: Posts about sensitive issues such as politics, sex, and suicide can be flagged and removed by algorithms, resulting in ‘seg’ and ‘mascara’ for sex and ‘unfortunate’ for death. Creative misspellings and stand-ins will be used, such as in the case of “alive”, “treasury”. for homophobia. Kurtz points out that there is a history of cryptographic responses to the ban. For example, 19th-century British Cockney prosodic slang and the allegorical language “Isopian” used to evade censorship in Tsarist Russia.
Algorithms aren’t the only ones that can’t recognize code. Euphemisms and misspellings are especially prevalent in marginalized communities. However, hidden language can also escape the human eye and can be potentially misleading online. In February, celebrity Julia Fox went public with an awkward exchange with a sexual assault victim after she misunderstood a post about “mascara” and inappropriately responded to what she thought was a discussion about makeup. I had to apologize on the spot.
thank you for reading!
Please let us know what you think. Please email your comments and suggestions to dealbook@nytimes.com.
