دانلود کتاب ابر آژانس – چه چیزی می تواند با آینده هوش مصنوعی ما درست پیش برود

Throughout history, new technologies have regularly sparked visions of impending dehumanization and societal collapse. The printing press, the power loom, the telephone, the camera, and the automobile all faced significant skepticism and sometimes even violent opposition on their way to becoming mainstays of modern living. Fifteenth-century doom-mongers argued that the printing press would dramatically destabilize society by enabling heresy and misinformation, and by undermining the authority of the clergy and scholars. The telephone was characterized as a device that could displace the intimacy of in-person visits and also make friends too transparent to one another. In the early decades of the car’s ascent, critics claimed it was destroying family life, with unmarried men choosing to save up for Model Ts instead of getting married and having kids, and married men resorting to divorce to escape the pressures of consumption that cars helped create. This same kind of doom and gloom was applied to society-wide automation in the 1950s, when increasingly sophisticated machines were dramatically impacting factories and office buildings alike, with everyone from bakers, meatcutters, autoworkers, and U.S. Census Bureau statisticians seeing their overall numbers dwindle. In 1961, Time magazine reported that labor experts believed that without intervention from business interests, unions, and the government, automation would continue to grow the “permanently unemployed.”3 By the mid-1960s, congressional subcommittees were regularly holding hearings regarding the mainframe computer’s potential threat to privacy, free will, and the average citizen’s capacity to make a life of their own choosing. Today, U.S. unemployment rates are lower than they were in 1961. The average U.S. citizen lives in a world where PCs, the internet, and smartphones have ushered in a new age of individualism and selfdetermination rather than crushing authoritarian compliance or the end of humanity. But with the emergence and ongoing evolution of highly capable AIs, it’s not just that familiar fears about technology persist; they’re growing. Even among AI developers, some believe that future instances of superintelligent AIs could represent an extinction-level threat to humanity. Others point out that, at the very least, humans acting with malicious intent will be able to use AIs to create catastrophic damage well before the machines themselves wage unilateral war against humanity. Additional concerns include massive job displacement, total human obsolescence, and a world where a tiny cabal of techno-elites capture whatever benefits, if any, AI enables. The doomsday warnings are different this time, these observers insist, because the technology itself is different this time. AI can already simulate core aspects of human intelligence. Many researchers believe it will soon attain the capacity to act with complete and extremely capable autonomy, in ways that aren’t aligned with human values or intentions. Robots and other kinds of highly intelligent systems have long existed in sci-fi novels, comic books, and movies as our dark doppelgangers and adversaries. So as today’s state-of-the-art AIs hold forth like benevolent but coolly rational grad students, it’s only natural to see foreshadowing of HAL from 2001: A Space Odyssey, or the Borg from Star Trek, or, in a less selfaware and more overtly menacing form, The Terminator’s relentless killer robot. These narratives have shaped our worst visions of the future for a long, long time. But are they the right narratives? The future is notoriously hard to foresee accurately—for pessimists and optimists alike. We didn’t get the permanent mass unemployment that labor experts in the early 1960s anticipated; nor did we get The Jetsons and its flying cars—at least not yet. As hard as it may be to accurately predict the future, it’s even harder to stop it. The world keeps changing. Simply trying to stop history by entrenching the status quo—through prohibitions, pauses, and other efforts to micro-manage who gets to do what—is not going to help us humans meet either the challenges or the opportunities that AI presents.