2 ms·
Hm, ok what if we introduce time limits and also only allow the computer to use as much energy as the human brain does?
by pazimzadeh 4y ago
Hm, ok what if we introduce time limits and also only allow the computer to use as much energy as the human brain does?
- cayleyh 4y agoThe results might surprise you: An new iPhone can run the latest chess engines at a level that should beat all but the absolute best players (https://www.hiarcs.com/iphone-chess.html https://www.hiarcs.com/iphone-chess.html). Apple doesn't publish power stats for their A-series chips that I could find, and I couldn't find a direct comparison of power draw running a chess engine, but Anandtec's numbers running synthetic benchmarks indicate between 3-5 watts when under single thread load: https://www.anandtech.com/show/16983/the-apple-a15-soc-performance-review-faster-more-efficient/2 https://www.anandtech.com/show/16983/the-apple-a15-soc-perfo.... Estimates for human brain energy consumption are vague, but generally accepted somewhere between 12-25 watts, with 20 being the most used average guess (https://www.quora.com/How-much-energy-does-a-human-brain-consume?share=1 https://www.quora.com/How-much-energy-does-a-human-brain-con..., https://www.scientificamerican.com/article/thinking-hard-calories/ https://www.scientificamerican.com/article/thinking-hard-cal..., https://www.munichre.com/topics-online/en/digitalisation/interview-henning-beck.html https://www.munichre.com/topics-online/en/digitalisation/int...) Given these numbers, not infeasible to assume that a computer restricted to the energy usage of the human brain would win every game when not playing Magnus?