law of accelerating returns
In plain wordsRay Kurzweil's claim that technology improves at an ever-faster rate, because each advance helps build the next, so progress speeds up exponentially rather than steadily.
Ray Kurzweil's thesis that progress feeds on itself: because each generation of technology helps build the next and the gains are reinvested, the pace of change in evolving systems, technology included, rises exponentially rather than holding steady. He argues that the price-performance of computing has followed a smooth exponential curve through five successive paradigms (electromechanical calculators, relays, vacuum tubes, discrete transistors and integrated circuits), with Moore's lawMoore's law, n.: Gordon Moore's prediction that chipmakers could pack twice as many transistors onto a chip roughly every two years — a pace that for decades made computers steadily faster and cheaper, and has slowed but not stopped. only the fifth, and that new paradigms take over as old ones saturate. From this he projected that computers would pass a valid Turing testTuring test, n.: Alan Turing's proposed test of machine intelligence: a judge holds typed conversations with a hidden person and a hidden computer, and the computer passes if the judge cannot consistently pick it out. by 2029 and that a technological singularitytechnological singularity, n.: A moment some people expect in the future when technology, usually AI smarter than us, starts changing the world so fast and so deeply that we cannot see past it. would arrive around 2045. Supporters cite the long-run growth of computation and, since about 2010, of AI training compute; critics object that an exponential trend in one metric does not imply exponential growth in capability or in technology generally, that the data are selected to fit the curve, and that most trends eventually saturate.
Kurzweil reads the scaling of training compute as one more instance of the law of accelerating returns.
Origin
Introduced by Ray Kurzweil in The Age of Spiritual Machines (1999) and set out in his essay 'The Law of Accelerating Returns' (2001); elaborated in The Singularity Is Near (2005) and The Singularity Is Nearer (2024).
Related
- technological singularitytechnological singularity, n.: A moment some people expect in the future when technology, usually AI smarter than us, starts changing the world so fast and so deeply that we cannot see past it.
- Ray KurzweilRay Kurzweil, prop. n.: An American inventor of reading machines and speech technology who became the best-known prophet of the 'singularity', predicting human-level AI by 2029.
- Moore's lawMoore's law, n.: Gordon Moore's prediction that chipmakers could pack twice as many transistors onto a chip roughly every two years — a pace that for decades made computers steadily faster and cheaper, and has slowed but not stopped.
- scaling lawsscaling laws, n.: Measured patterns showing that an AI model's error shrinks smoothly and predictably as the model is made bigger and given more data and computing power, so labs can forecast a giant model's performance from small, cheap experiments.
- AI timelinesAI timelines, n.: Forecasts of when AI will reach major milestones, such as matching humans at most work. People with short timelines expect this within years; others expect decades, or never.
- intelligence explosionintelligence explosion, n.: The idea that once AI gets good enough at improving AI, each smarter version could build a smarter one faster still, so abilities would snowball far beyond human level in a short time.
Sources
- Kurzweil, R. (2001). The Law of Accelerating Returns. KurzweilAI.net.
- Kurzweil, R. (1999). The Age of Spiritual Machines: When Computers Exceed Human Intelligence. Viking.
Mentioned in
Moore's lawMoore's law, n.: Gordon Moore's prediction that chipmakers could pack twice as many transistors onto a chip roughly every two years — a pace that for decades made computers steadily faster and cheaper, and has slowed but not stopped., technological singularitytechnological singularity, n.: A moment some people expect in the future when technology, usually AI smarter than us, starts changing the world so fast and so deeply that we cannot see past it., Ray KurzweilRay Kurzweil, prop. n.: An American inventor of reading machines and speech technology who became the best-known prophet of the 'singularity', predicting human-level AI by 2029.