Skynet Is (Almost) Here

Readers of CTRL+ATL+REVOLT and similar tales of A.I. vs. Mankind, you want to read this post at Global Guerillas. It's about how a next-gen program beat a Go Grandmaster three games straight, and fast. I'll quote the relevant parts below:

Instead, AlphaGo won this victory by learning how to play the game from scratch using this process:

• No assumptions. AlphaGo approached the game without any assumptions. This is called a model-free approach. This allows it to program itself from scratch, by building complex models human programmers can't understand/match.
  
• Big Data. It then learned the game by interacting with a database filled with 30 million games previously played by human beings. The ability to bootstrap a model from data removes almost all of the need for engineering and programming talent currently needed for big systems. That's huge.
  
• Big Sim (by the way, Big Sim will be as well known as Big Data in five years <-- heard it here first). Finally, it applied and honed that learning by playing itself on 50 computers night and day until it became good enough to play a human grandmaster.

That, right there, should scare the living shit out of you. Global Guerillas being what it is, it lays out the implications:

• This technology is going to cut through the global economy like a hot knife through butter. It learns fast and largely on its own. It's widely applicable. It doesn't only master what it has seen, it can innovate. For example: some of the unheard of moves made by AlphaGo were considered "beautiful" by the Grandmaster it beat.
  
• Limited AGI (deep learning in particular) will have the ability to do nearly any job currently being done by human beings -- from lawyers to judges, nurses to doctors, driving to construction -- potentially at a grandmaster's level of capability. This makes it a buzzsaw.
  
• Very few people (and I mean very few) will be able to stay ahead of the limited AGI buzzsaw. It learns so quickly, the fate of people stranded in former factory towns gutted by "free trade" is likely to be the fate of the highest paid technorati. They simply don't have the capacity to learn fast enough or be creative enough to stay ahead of it.

Well, shit.

I've been consistent for a few years now that the automation trend will so severely screw the market for human employment that we may well face a threat to Civilization just from the consequences alone. It's not just the unskilled cohort doing menial labor; they're just the the first to get the squeeze. What this technology means is massive deskilling of knowledge-based work, especially anything computer-related.

IT is soon to be taken out of human hands. As soon as robotics are equal to human capability, Medicine will go from partial to full deskilling; the AIs will first handle all knowledge-based work, and then do the physical stuff. The AIs will take over the omnipresent surveilance networks and law enforcement will soon go out of human hands as a result. Soon so will jurisprudence and regulation; these two are knowledge-based work, and when the AIs see that they're better at it than we are they will just usurp it for the greater good.

So, what will we do when we no longer can work due to be unfit for purpose and surplus to requirements? When the state is run by the machines we created, running on cold logic?

That's what Hawking warned us about. That's what SF writers for generations have warned us about, since the very first play that coined the word "robot". Christ, folks, can we please stop creating our own demises? If we're lucky, a Bulterian Jihad will be the least bothersome of responses.