Could Facebook’s new Timeline lead to Frankenstein’s software?

by Andrew Scott

Bringing people back from the dead may sound like a flight of fancy, but back in 1999, using Moore’s law, self-made technology mystic Ray Kurzweil made the claim that by 2020 the computing power of a $1000 PC would match the processing speed and capacity of the human brain. Now I’m not sure how much stock I’d put in some of Ray’s other predictions, but this one seems fairly reasonable.

Add to this prediction, what CEO of Digital Sky Technologies and Facebook investor Yuri Milner has said about Facebook’s potential, to be a breeding ground for AI programs, and you have a heady mix of converging technologies that could, potentially, lead to Mary Shelly like consequences.

What the hell am I getting at? Well follow me down the rabbit hole if you will.

We all know that Facebook is rolling out its new profile page called Timeline, on which users will be encouraged to make a scrapbook of their lives from birth, to the present day and beyond. Is there really much of a difference between this online Timeline and the memories we have stored organically in our brains?

If you are still prepared to follow me, is it not conceivable that Facebook’s Timeline could form the basis of the memories of some future AI software, essentially allowing us to bring people back from the dead and at the same time, perhaps, giving us a way to live forever?

This might sound like an episode of Caprica, but consider this, IBM has already built a prototype processor capable of parallel processing, working in much the same way as our human brains do; and the same company has made a super computer, Watson, intelligent enough to win on the TV show Jeopardy – though this might not impress some people much – the significance of this achievement is astonishing.

A third technology I’d like to bring into the mix is evolutionary computation. When I was at school I can remember my IT teacher saying – in a very slow voice – “computers are s-t-u-p-i-d you’ve got to tell them what to do”. However Stephanie Forrest and her collaborators at The Santé Fe Institute may not be that far away from creating software that can write its own code through trial and error.

Moving on, imagine the internet as the primordial soup of data through which simplistic self-replicating programs could swill about rewriting their own code at random, beneficial lines of code being kept, and less beneficial lines of code being discarded. It’s not too hard to imagine that in a very short period of time, very novel programs will begin to emerge, well beyond the programming powers of any group of IBM whizzes.

So when you add together all the raw ingredients; IBM’s processing power, Facebook’s Timeline and Stephanie Forrest’s self-evolving software; you have a fairly compelling argument that we’re not that far away at all, from screaming “IT’S ALIVE!”

Leave a Reply