fast takeoff
In plain wordsA scenario in which AI goes from about human level to far beyond it very quickly, in days, weeks or a few months, leaving people little time to react.
A takeofftakeoff, n.: The stretch of time between AI reaching roughly human level and AI becoming vastly superhuman. People argue over whether it would take decades, years, months or just days. in which the transition from roughly human-level AI to superintelligencesuperintelligence, n.: A mind, usually imagined as an AI, that would leave even the cleverest people far behind in nearly every kind of thinking, not just in one game or skill. None is known to exist. is short compared with the time human institutions need to respond. Nick BostromNick Bostrom, prop. n.: A Swedish philosopher whose 2014 book 'Superintelligence' argued that AI smarter than humans could be catastrophic unless its goals are carefully controlled, and helped make AI risk a mainstream concern. defined a fast takeoff as one occurring over minutes, hours or days, with a moderate takeoff taking months or years; later usage often stretches the threshold to weeks or months. The usual mechanisms are recursive self-improvementrecursive self-improvement, n.: An AI improving its own design or training, so that each improved version is better at making the next improvement. It is the engine usually imagined behind an intelligence explosion. and a large compute overhangcompute overhang, n.: A situation in which far more computing power is available than AI currently uses, so that a new method or a decision to scale up could suddenly produce a big jump in ability.: once a system can improve itself or exploit abundant idle compute, capability growth outpaces anyone's ability to notice and intervene. A fast takeoff implies that safety measures must largely be in place beforehand, since there may be no chance to learn from intermediate systems, and makes it more likely that one project gains a decisive strategic advantagedecisive strategic advantage, n.: A lead in technology and power so large that whoever holds it, whether a country, a company or an AI, could impose its will on the whole world.. Paul Christiano defined it negatively, as any takeoff that is not a slow takeoffslow takeoff, n.: A scenario in which AI becomes superhuman step by step, over years or decades, as it spreads through the economy, giving society time to spot problems and respond. in his economic sense.
In a fast takeoff, the first warning sign might also be the last.
Origin
'Hard takeoff' appears in Eliezer Yudkowsky's Creating Friendly AI (2001), its earliest use not established, and was defended in his 2008 debate with Robin Hanson; 'fast takeoff' became standard after Nick Bostrom's Superintelligence (2014), whose chapter on the kinetics of an intelligence explosion distinguished slow, moderate and fast transitions.
Usage note
Often conflated with short 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.; the two are logically independent.
Related
- takeofftakeoff, n.: The stretch of time between AI reaching roughly human level and AI becoming vastly superhuman. People argue over whether it would take decades, years, months or just days.
- slow takeoffslow takeoff, n.: A scenario in which AI becomes superhuman step by step, over years or decades, as it spreads through the economy, giving society time to spot problems and respond.
- FOOMFOOM, n.: Slang for a sudden, explosive jump in an AI's abilities, the AI going 'foom' and racing from about human level to vastly superhuman in a very short time.
- 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.
- recursive self-improvementrecursive self-improvement, n.: An AI improving its own design or training, so that each improved version is better at making the next improvement. It is the engine usually imagined behind an intelligence explosion.
- decisive strategic advantagedecisive strategic advantage, n.: A lead in technology and power so large that whoever holds it, whether a country, a company or an AI, could impose its will on the whole world.
- sharp left turnsharp left turn, n.: A feared moment when an AI's abilities suddenly leap far beyond its training while its good behavior fails to keep up, leaving it powerful but no longer aligned.
- slow takeoffslow takeoff, n.: A scenario in which AI becomes superhuman step by step, over years or decades, as it spreads through the economy, giving society time to spot problems and respond.
Mentioned in
slow takeoffslow takeoff, n.: A scenario in which AI becomes superhuman step by step, over years or decades, as it spreads through the economy, giving society time to spot problems and respond., takeofftakeoff, n.: The stretch of time between AI reaching roughly human level and AI becoming vastly superhuman. People argue over whether it would take decades, years, months or just days., compute overhangcompute overhang, n.: A situation in which far more computing power is available than AI currently uses, so that a new method or a decision to scale up could suddenly produce a big jump in ability., decisive strategic advantagedecisive strategic advantage, n.: A lead in technology and power so large that whoever holds it, whether a country, a company or an AI, could impose its will on the whole world., FOOMFOOM, n.: Slang for a sudden, explosive jump in an AI's abilities, the AI going 'foom' and racing from about human level to vastly superhuman in a very short time., sharp left turnsharp left turn, n.: A feared moment when an AI's abilities suddenly leap far beyond its training while its good behavior fails to keep up, leaving it powerful but no longer aligned., warning shotwarning shot, n.: An AI accident or near-miss serious enough to alarm the world about AI's dangers, but not so serious that it cannot be recovered from, which could prompt action before a real catastrophe.