Welcome back to In the Loop, TIME\u2019s twice-weekly newsletter about the world of AI. If you’re reading this in your browser, you can subscribe to have the next one delivered straight to your inbox.<\/p>\n
Subscribe to In the Loop<\/a><\/p>\n What to Know: The AI social contract<\/p>\n At a lakefront venue in Sweden earlier this month, 18 individuals from OpenAI, Google DeepMind, the U.K. AI Security Institute, the OECD, and other groups gathered for an invite-only summit. On the agenda: arriving at an understanding of the likely ways that advanced AI will impact the \u201csocial contract\u201d between working people, governments, and corporations.<\/p>\n Top AI CEOs like DeepMind\u2019s Demis Hassabis and OpenAI\u2019s Sam Altman have recently been urging<\/a> academics and governments to grapple with this issue more deeply, to better prepare the world for what they expect will be a highly disruptive economic shock. So, every day for a week\u2014in breakout rooms and in a nightly communal sauna\u2014these 18 experts hashed out a picture of what economic shocks might be coming down the track\u2026 and what to do about them.<\/p>\n Bad news \u2014 One outcome of the so-called \u201cAGI social contract summit\u201d was a list of four draft statements, according to the summit\u2019s organizers. These statements have not previously been reported. They paint a grim picture of where the world could be headed, absent significant interventions by governments and societies. \u201cAI is likely to exacerbate increasing wealth and income inequality within countries, worsening economic conditions for many working and middle-class people and families,\u201d the first reads. \u201cAI will increase inequality between countries that have access to AI infrastructure and those that don\u2019t\u2014both in terms of access to benefits as well as ability to respond to shocks,\u201d says the second. \u201cWithout intervention, AI-enabled inequalities may lead to the political dominance of wealthy individuals and corporations, eroding democratic institutions and increasing levels of political dissatisfaction,\u201d the third says. And the fourth: \u201cThe encroachment of AI systems and the erosion of the value of labor could lead to the increasing disempowerment of most humans, causing a degradation in individual well-being and purpose.\u201d<\/p>\n Human disempowerment \u2014 Attendees at the summit agreed that the existing social contract\u2014in which people receive security and a stake in society in return for their labor\u2014is in trouble due to AI, says Deric Cheng, the event\u2019s organizer, who serves as Director of Research at the Windfall Trust, a non-profit founded this year to grapple with these issues. \u201cWe’re essentially worried that labor will be disempowered relative to corporations, and also to some degree that governments might be disempowered relative to corporations,\u201d Cheng says. \u201cThe obvious result of lower labor power is decreased real wages.\u201d This view holds that people in wealthy democracies enjoy a high standard of living not due to their rights enshrined on paper\u2014but due to their ability to withhold their labor. Remove labor from that equation, and standards of living are vulnerable to going down, even if overall GDP or productivity statistics rise.<\/p>\n Ways forward \u2014 Without intervention by governments, attendees agreed, the default path of advanced AI would likely result in bad economic outcomes for the average person. But fortunately, they also identified several possible actions that governments could take to push things in a better direction, Cheng says. For example: developing new institutions, in the vein of the IMF, to ensure that wealth derived from AI is distributed globally, rather than within the one or two powerful countries where AI companies are located. States could also run pilots today, Cheng says, for policies like basic income and reduced working weeks, to gather evidence about what kinds of safety nets are effective. <\/p>\n Google DeepMind declined to comment on the statements that arose from the summit. OpenAI did not respond to a request for comment. After the publication of this article, Cheng said in a message that all attendees were present in a personal capacity, not an official capacity, and that the statements do not reflect the views of their organizations. He also added that some attendees disagree with some of the draft statements, which he had previously described as a “consensus” of the group.<\/p>\n If you have a minute, please take our quick survey to help us better understand who you are and which AI topics interest you most.<\/p>\n Take the survey<\/a><\/p>\n Who to Know: U.S. District Judge Amit Mehta<\/p>\n Last year, Federal District Judge Amit Mehta ruled that Google had illegally maintained a monopoly over online search and ads. This week, he is expected to announce the court\u2019s decision on what to do about it\u2014a ruling that could range from making Google share data with rivals, to forcing a breakup of the search giant itself.<\/p>\n Payments to rivals \u2014 The U.S. Department of Justice\u2019s case against Google revolved around the multibillion dollar yearly payments that Google made to Apple in order to secure Google as the default search engine on iPhones. Observers expect the court to, at a minimum, place limits on these kinds of payments, which Mehta ruled were anticompetitive.<\/p>\n Spinning off Chrome \u2014 Another possibility is that Mehta could order Google to sell Chrome, the most popular browser in the world, with a 67% market share. Chrome allows Google to collect intricate data about users\u2019 browsing patterns that shore up its dominance of the search and ad space. Any of Google\u2019s competitors would no doubt jump at the chance to buy the world\u2019s top browser, given the opportunity it affords to point users toward their LLM of choice. <\/p>\n Sharing user data \u2014 The data that Google collects on its users is part of the secret sauce of its search engine. Mehta could rule that Google must share this data with competitors\u2014perhaps in an anonymized form, to ward off accusations of privacy violations. <\/p>\n AI in Action<\/p>\n The public trusts AI chatbots more than companies or community leaders, according to polling<\/a> of users in 68 countries carried out by the Collective Intelligence Project.<\/p>\n More than half (56.6%) of people trust AI chatbots, the polling found. That\u2019s higher than the AI companies that make them (34.6%) or even faith and community leaders (44.2%). <\/p>\n More than one in 10 people (14.9%) are using AI for emotional support on a daily basis, the survey found. And 30% of people have \u201cat some point thought their AI chatbot might be self-aware.\u201d <\/p>\n And 56% of people polled said that the proliferation of AI across society was likely to worsen access to good jobs.<\/p>\n As always, if you have an interesting story of AI in Action, we\u2019d love to hear it. Email us at: intheloop@time.com<\/a><\/p>\n What We\u2019re Reading<\/p>\n