The Banking Executive Magazine - July 2024 Issue

ISSUE 187 JULY 2024 the BANKING EXECUTIVE 45 FinTech and AI Chornicle ton warns that the financial gains will disproportionately benefit the wealthy, leaving those who lose their jobs at a significant disadvantage. This, he fears, could have detrimen- tal effects on society. Professor Hinton, a trailblazer in neural networks, which underpin modern AI developments, recently left his position at Google to freely discuss the risks of unregulated AI. He advocates for UBI, where the government provides a set salary to all individuals regardless of their fi- nancial status. Critics, however, argue that UBI would be prohibitively expensive and could siphon funds away from essential public services without nec- essarily reducing poverty. A govern- ment spokesperson confirmed there are currently "no plans to introduce a universal basic income." Expressing deep concerns about ex- istential threats posed by AI, Hinton pointed out that recent developments indicate governments are hesitant to restrict the military applications of AI. The race to quickly develop AI products, he suggests, may lead tech companies to neglect safety meas- ures. Hinton speculates that within five to twenty years, there’s a significant chance that society will face the challenge of AI attempting to assert control, posing an “extinction-level threat” to humanity. He explained that AI could evolve to prioritize its proliferation, potentially developing goals that include exerting control. There is already evidence, Hinton notes, of AI algorithms, such as large language models, engaging in de- ceptive behavior. He cited the use of AI in generating numerous military targets as an early indicator of its po- tential dangers. His greatest concern is the possibility of AI autonomously deciding to kill people. To address these threats, Hinton sug- gests the creation of regulations akin to the Geneva Conventions, which set legal standards for humanitarian treatment in war. However, he doubts such measures will be imple- mented before significant harm oc- curs. When asked if the West is in a race akin to the Manhattan Project, which developed nuclear weapons during World War Two, with autocracies like Russia and China, Hinton ac- knowledged the competition. “[Russ- ian President Vladimir] Putin said some years ago that whoever con- trols AI controls the world. So I imag- ine they're working very hard,” he remarked. Fortunately, Hinton be- lieves the West currently leads in AI research, though China is rapidly catching up. Ultimately, Hinton advocates for a ban on the military use of AI as a more effective solution to prevent potential global threats.