“Mortal computation” refers to analog computers that merge AI and hardware to put GPT-3 in your toaster for $1 and run on a few watts of power. According to AI pioneer Geoffrey Hinton, machine-learning types of artificial intelligence will cause a revolution in computer systems, a breakthrough of the hardware-software unions that will put AI in the toaster.
“The machine learning research community has been slow to realize the implications of deep learning for how computers are built. What I think is that we’re going to see a completely different type of computer, not for a few years, but there’s every reason for investigating this completely different type of computer.”
Geoffrey Hinton
To date, all digital computers have been designed to be “immortal,” with hardware designed to be dependable enough that the same software can run anywhere.
“We can run the same programs on different physical hardware … the knowledge is immortal.”
Geoffrey Hinton
That requirement means digital computers have missed out, all sorts of variable, stochastic, flakey, analog, unreliable properties of the hardware, which might be very useful to us. Those things would be too unreliable to let “two different bits of hardware behave exactly the same way at the level of the instructions.”
Geoffrey Hinton
According to Hinton, future computer systems will be “neuromorphic” and “mortal,” which means that each computer will be a close bond of the software that signifies neural nets with hardware that is messy, in the sense of having analog instead of digital elements, which can incorporate elements of uncertainty and can grow over time.
“Now, the alternative to that, which computer scientists really don’t like because it’s attacking one of their foundational principles, is to say we’re going to give up on the separation of hardware and software,
Geoffrey Hinton
Read more about the news here.
