TED - Don't fear intelligent machines. Work with them | Garry Kasparov

Short Summary

Main recommendation: Don't fear intelligent machines; instead, work with them in a human-plus-machine partnership where people supply intuition, purpose, and judgment and machines supply calculation, memory, and speed. Kasparov argues that the "man versus machine" framing is misleading and that humans should keep pushing themselves toward harder challenges rather than retreating from technological progress.

Key Points

• Kasparov's experience with Deep Blue in 1997 showed that machines can reach grandmaster-level chess through brute-force calculation (about 200 million positions per second), without gaining the insight into human intelligence that early computer scientists hoped for.
• His "Advanced Chess" experiments (from 1998) initially failed to combine human and machine skill well, but a 2005 freestyle tournament was won by two amateur players running three ordinary PCs, showing that a better process matters more than raw strength.
• Human-plus-machine collaboration is already common: people use online translation and correct it, and the same model is spreading into investing, medical diagnosis, and security analysis, where machines do most of the data work and humans make the final decisions.
• Despite doomsayer predictions, chess has grown more popular even though free apps now outplay Deep Blue, suggesting people value human competition and meaning regardless of machine ability.
• Kasparov argues that technology removes difficulty and uncertainty, so humans must seek harder and more uncertain challenges rather than slow progress.

Details Worth Noting

• Concrete instructions: Kasparov's practical guidance is to combine the machine's strengths (calculation, data crunching, probability estimates) with human strengths (intuition, strategy, experience, purpose), and to treat human corrections as training for the machine. Specifically, when a task is machine-assisted, let the machine get the analysis to roughly 80 to 90 percent, then have a human check and make the decision. He also advises facing fears about technology directly, and asking not what machines can do today but what they still cannot do, then working with new intelligent machines toward those goals. The video does not give numeric training schedules, tools, or step-by-step setups.
• The 1985 Hamburg exhibition: Kasparov, then 22, won all games against 32 chess machines at once, before becoming world champion by beating Anatoli Karpov that year.
• Deep Blue match history: Kasparov won the first match in Philadelphia, then lost the rematch in New York the following year, and he noted that the machine's triumph was also a triumph for its human creators, Feng-hsiung Hsu, Murray Campbell, and their team.
• Kasparov's caveat on safety: he said he would not send his children to school in a self-driving car at 90 percent accuracy, and that even 99 percent may not be enough without further breakthroughs.
• He also cited John Henry, the 19th-century steel-driving folk legend, as an example of the long-running human-versus-technology narrative.

Bottom Line

Kasparov's core message is that fear of intelligent machines comes from framing progress as a contest, while the better path is collaboration plus ambition: let machines handle what they do well, keep humans responsible for purpose and judgment, and aim for harder problems rather than complacency.