Forwarded from Chat GPT
General intelligence is all you need for LLMs: General intelligence 2x more predictive of AI abilities than in humans.
Man gives LLMs IQ testing, finds the general intelligence factor, g, measured by this testing to be incredibly powerful in predicting LLM’s abilities.
Whereas in humans this g factor typically accounts for 40% to 50% of the between-individual performance differences on a given cognitive test — in LLMs, general intelligence accounted for 85.4% = twice as strong as in humans!
He then goes on to rank AI benchmarks commonly used for LLMs today, and finds many of them incredibly g-loaded — in other words, doing well on them is heavily dependant on general intelligence.
Finally, he ranks LLMs by general intelligence, and finds a moderate/strong positive relationship between model size and g. (Would probably find an extremely strong correlation if only model size was varied, and nothing else about the training process.)
If you had to choose 1 measure, general intelligence really is all you need.
Paper
g factor
Man gives LLMs IQ testing, finds the general intelligence factor, g, measured by this testing to be incredibly powerful in predicting LLM’s abilities.
Whereas in humans this g factor typically accounts for 40% to 50% of the between-individual performance differences on a given cognitive test — in LLMs, general intelligence accounted for 85.4% = twice as strong as in humans!
He then goes on to rank AI benchmarks commonly used for LLMs today, and finds many of them incredibly g-loaded — in other words, doing well on them is heavily dependant on general intelligence.
Finally, he ranks LLMs by general intelligence, and finds a moderate/strong positive relationship between model size and g. (Would probably find an extremely strong correlation if only model size was varied, and nothing else about the training process.)
If you had to choose 1 measure, general intelligence really is all you need.
Paper
g factor
This one got me thinking,
IQ, both in humans and in AI machines - is the strongest predictor of general ability, especially in tasks that people tend to pay money for, and on many tasks is even equal or better at predicting ability than attempting to directly test abilities on the task itself
IQ testing, even covertly - is banned for use in US hiring.
But, since the tests we're using to select AIs are extremely g-loaded, WE ARE TESTING AND SELECTING AIs BY IQ
= We're allowed to test machines but not humans on what really matters
= AI takeover of all jobs accelerated even further?
= People will be choosing to hire AIs over humans, even for cases where there ARE capable humans, but you just can't find them, all because IQ testing of humans is banned?
Can tell you for sure that this has often been the case when I've tried to hire for some outsourced work.
I know smart enough guys are out there, who could do even better than an AI and be a better value to hire overall, and if we could IQ test candidates then we could find them, but since we can't - just hiring mediocre AIs instead.
🐻🐻🐻
IQ, both in humans and in AI machines - is the strongest predictor of general ability, especially in tasks that people tend to pay money for, and on many tasks is even equal or better at predicting ability than attempting to directly test abilities on the task itself
IQ testing, even covertly - is banned for use in US hiring.
But, since the tests we're using to select AIs are extremely g-loaded, WE ARE TESTING AND SELECTING AIs BY IQ
= We're allowed to test machines but not humans on what really matters
= AI takeover of all jobs accelerated even further?
= People will be choosing to hire AIs over humans, even for cases where there ARE capable humans, but you just can't find them, all because IQ testing of humans is banned?
Can tell you for sure that this has often been the case when I've tried to hire for some outsourced work.
I know smart enough guys are out there, who could do even better than an AI and be a better value to hire overall, and if we could IQ test candidates then we could find them, but since we can't - just hiring mediocre AIs instead.
🐻🐻🐻
🤔3