It depends. A lot of narrow AI was developed in academic and other research settings for not much money. If what is needed is largely a set of successful insights and their integration this is likely not so tractable from merely large sums of money. If the path is brain emulation that is indeed a rather expensive proposition.
2) Machines need to design machines.
There are two aspects. Hardware and software. I had comment limits!
The last sentence is especially important. If the first super-intelligent machine is indoctrinated in corporate culture, then environmental destruction doesn't seem far off. If they have nano, then the grey goo scenario is a disturbing possibility.
Assumptions:
1) AGI will take a lot of money to develop
It depends. A lot of narrow AI was developed in academic and other research settings for not much money. If what is needed is largely a set of successful insights and their integration this is likely not so tractable from merely large sums of money. If the path is brain emulation that is indeed a rather expensive proposition.
2) Machines need to design machines.
There are two aspects. Hardware and software. I had comment limits!
serendipityseraph 8 months ago
The last sentence is especially important. If the first super-intelligent machine is indoctrinated in corporate culture, then environmental destruction doesn't seem far off. If they have nano, then the grey goo scenario is a disturbing possibility.
wonderist 3 years ago