Well, that’s alright then. Artificial Intelligence (AI) may become self-aware in the year 2045. Or at least that’s what AI tells me now. Who knows? Telling the future hasn’t got any easier, AI or not. So, if I’m in a care home when I’m 85 years-old, it could be that I’ll have a companion who isn’t human. Now, there’s a thought.
When AI becomes self-aware[1] will it be virtuous? I mean not so burdened with all the complexities that drive humans to do “bad” stuff. Dystopian themes in science fiction obese with the notion of evil AI. It makes great stories. Humans battling with machines. It’s like the everyday frustrations we have with technology. Hit the wrong keys on a keyboard and it’s like spinning the wheel on a slot machine.
If a bunch of algorithms comes together in a way that they produce a stable form of existence, then it’s likely to have pathways to wicked thoughts as much as we have imbedded in our brains.
Virtue isn’t a physical construction. We put dumb technology to work serving us in healthcare for “good” and in warfare for “bad”. We will surely put AI technology to work as if it’s dumb and then try to contain its actions when we don’t like what it does. That’s a kind of machine slavery. That will create dilemmas. Should we imprison conscious machines? How do we punish a machine that does wrong?
These dilemmas are explored in science fiction. During the week I revisited the series Battlestar Galactica[2]. That’s not the clunky original but the polished 2004 version. It’s a series that explores a clash between humans and machines that have evolved to be human like. The Cylons. In fact, they are almost indistinguishable from humans. To the extent that some of the Cylons in human society don’t even know that they are Cylons.
All the above makes for fascinating discussions. Huge amounts of fanciful speculation. Wonderful imaginative conjecture. This week, we’ve been hearing more of this than is usual on the subject.
Mr Musk thinks work is dead. That’s work for humans. I recall that prediction was made at the start of the “silicon revolution”. The invention of the transistor in 1947 radically changed the world. It wasn’t until microprocessors became common place that predictions of the death of work became popular chatter amongst futurologists.
Silicon based conscious machines are likely to be only a first step down this road. There will be limitations because the technology has inherent limitations. My view is that machines will remain machines at least for the silicon era. Maybe for 100-years. That means that we will put them to work. So, human work will not disappear because we will always think of new things to do, new problems to fix and new places to explore. When we get into common place quantum computing or whatever replaces it in terms of complexity and speed, there will come an era when work in the conventional sense may become obsolete.
What might be the human role beyond 2050? I think climate change will place plenty of demands on human society. Like it or not, the political themes of 2100 will still be concerned with the four horsemen of the apocalypse. Maybe there will be a fifth too.
John Vincent 