AI ‘magically’ won’t solve problems, humans can’t already solve, says Bill Gates

Bill Gates recently shared his insights on Artificial Intelligence (AI) with Dax Shepard on the Armchair Expert podcast. Gates, a known advocate for AI development, acknowledges both the immense potential and the current limitations of this technology, according to a report by Business Insider. Gates sees AI as a game-changer in fields like pharmaceuticals and agriculture, areas his foundation actively supports.However, he's not entirely impressed by all its feats, particularly AI's struggle with complex tasks like Sudoku due to limitations in recursive reasoning.Gates emphasises the importance of data quality in AI development. "Good AI" requires "data that embodies the expertise," he says. This means AI excels at tasks with well-defined datasets but struggles with subjective or qualitative areas like human happiness or relationships. “If it's a problem that humans are not good at dealing with, then present techniques don't create some novel approach,” Gates said.While Gates sees potential for AI in mental health support, he stresses the need for careful implementation. “I actually think in mental counselling AIs will play a role, but boy, we're going to have to be very careful about that,” Gates said. “That's going to require a lot of work that has not been done yet,” he warns.Gates also highlights surprising shortcomings in complex math, where AI can't readily admit mistakes. "It doesn't know to check its answers," he explains, citing an example of AI attributing a wrong Sudoku answer to a "mistype" rather than an error.Despite these limitations, Gates remains optimistic. He sees the potential benefits as too significant to halt progress. He acknowledges differing opinions on AI development but doubts a global agreement to slow it down. "Maybe we'll have a world where some parts embrace AI fully, while others are more cautious," he suggests.