When we reach artificial general intelligence (and estimates say we will by 2030) it won’t just mean computers have human-level general intelligence.
Because even if a computer with AGI had the same levels of intelligence and computational power as a human, the AGI would still have significant advantages over us.
Informal estimates place the brain’s neural firing rates in the <1-200 Hz range.
Today’s top microprocessors (which will be even faster when we reach AGI) run in the range of 3-4 GHz, over 10 million times faster than our neurons.
A human’s nerve impulse can move in the range of 80-120 m/s.
A computer can communicate optically at the speed of light (299,792,458 m/s).
2. Size and storage
Our skulls limit the brain in its size and expansion.
Computers can expand to allow for a larger RAM (working memory) and hard drive storage (long-term memory) with greater precision.
3. Reliability and durability
Humans and their brains fatigue easily while a computer can run nonstop at peak performance.
Computers are also easier to repair and replace than biological neurons if they deteriorate.
Computers can receive updates and fixes quickly, unlike the human brain.
Humans excel against all other species on the planet at accumulating collective intelligence.
This is due to our ability to invent languages, form civilizations and develop technology to spread information (like the printing press and the internet).
But computers will be better than humans at this.
A network of AGI-caliber computers would connect in a way that allows them to instantly sync with each other once one learns a new piece of information.
AGI computers would also cooperate with each other without dissent as there will be no opinions or self-motivations, like humans have, to restrict them.
Source: Bostrom, Nick (2014). Superintelligence: Paths, Dangers, Strategies. Oxford: Oxford University Press