The Coming Technological Singularity (1993) | Hacker News Reader

The Coming Technological Singularity (1993)

www-rohan.sdsu.edu

15 pointsdoener8 comments

This gets posted from time to time. I highly recommend Vernor Vinge's A Fire Upon the Deep. One of my favorite SciFi novels of all time. When I read it (90s), before the www was in wide spread adoption and usenet was king, there are some very funny usenet riffs in the book. Anyone that was alive and aware of early usenet will appreciate it. My favorite Vinge usenet header: Translation-Path.

I've tried so many times to read that book, but I can never get past the packs of communal-brain dogs.

Indeed, it's essentially two books in one, and that half of it was always a slog to get through.

Marooned in Realtime is better for getting the Singularity across, although you really need to read The Peace War first; this collection has both plus a bridging short: http://www.amazon.com/Across-Realtime-Vernor-Vinge/dp/185798...

> Within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended.

This was written in 1993 so 30 years makes it 2023. That may not be too far off.

Fortune telling is a dangerous business.

It would be interesting to know if Vernor Vinge stands by his prediction and still thinks that the end of the human era is only 7 years away.

To be fair, he admits in the article that it might never happen. It was just an extrapolation based in a best guess of the computational complexity of the human brain and advances in computer technology. But it was still just an estimate.

It was also based on an implicit assumption that once we had computers equivalent in power to the brain, that shortly afterwards we would have computers as intelligent as a human brain. But that assumes that programming an AI, or replicating the architecture of the brain, is essentially a trivial task given the required hardware. I think it's pretty clear now that this isn't true. We have very little idea how the brain really works. Even now estimates of the computational power of neurons or cortical columns vary enormously and are essentially guesses.

> replicating the architecture of the brain, is essentially a trivial task given the required hardware. I think it's pretty clear now that this isn't true.

I think it will be like airplanes. We don't flap our wings but we fly.

In Rainbows End, written in 2006, he has a mid 2020s setting that seems to be on the cusp of the Singularity.

Reply on news.ycombinator.com