Ai. Whats that all about.
My youtube feed is overwhelmed by articles on Ai. Artificial Intelligence.
It's probably because I have shown some interest in the topic, or that the Robot Overlords don't think much of my ability to survive without some help from Ai.
To define Ai is to understand the premise that it is all built upon.
Moore's Law.
Every 18 months or so the number of transistors in a CPU will double with no corresponding rise in price. This doesn't actually explain much, so I looked at the Physicist Michio Kako's explanation.
He said he bought a birthday card for his daughter for a few cents. And when she opened it, the card sang "happy birthday!" to her.
When he examined what made this possible he found a tiny chip in the spine of the card. And established that that little chip had more computing power than Nasa's Apollo mission to the moon.
This was a really good illustration of Moore's law in motion.
So the theory is that if the computer's growth is exponential it will surely result in a machine who's level of complexity matches that of a human brain. And when that happens, 18 months later it would be twice as complex. Most in the industry agree that this won't happen for another 5 years, but will not take longer than 25 years from today. (2022 - 2042).
So what does this really mean?
If its a given that the machines of the future will surely be more complex than the human brain several things will happen.
- No human will have the ability to comprehend the device. (no system has the computational power to fully describe itself, let alone something more complex).
- The system will be able to learn and write its own software, and possibly improve on its design at an exponential pace.
- It will be able to improve the world at a level that would be beyond our ability to comprehend.
Most jobs would be lost to the emerging machines. Few professions would be safe from the incursions that Ai will make. Medical doctors, surgeons, Accountants, Lawyers would be particularly vulnerable to the advance of new automation and technology.
The fact that IBM Deep Blue was able to beat Gary Kasparov (Chess Grand Master) at the game that represented the pinnacle of human intelligence.
IBM's Watson was able to beat the world champ at Jeopardy (a TV game show that had come to represent the unique ability of the human mind to make intuitive connections between facts).
Finally when Googles Ai was able to beat the world champion "Go" player. (Go is a Chinese board game with billions of possibilities was thought to be impossible for machines)
These three examples served to reinforce the "fact" that the age of the Machines was upon us.
Self-driving cars are clearly ready, and this only reinforces the view that Ai is the future, and that this future is eminently upon us.
Some even go so far as to predict a world in which Humans would lose the position of the dominant species.
Some even go so far as to predict a world in which Humans would lose the position of the dominant species.
Some of the best minds in the world. Steven Hawking, Bill Gates, Elon Musk and many others have warned the scientific community about the risks associated with the development of Deep Ai. Some have even compared it to the development of the Atom Bomb in that it would represent a game changer in the shift of power it represented and in the wrong hands it could easily throw the world as we know it into chaos.
Humans might even end up competing with the new species to survive.
There is however a different, less pessimistic view.
There is however a different, less pessimistic view.
Its been a few years since I wrote this piece. Today is 9th Jan 2019.
The year after Google's AI beat the Go player, they created a new computer that was not schooled in the rules of Go and fed all the previous games of the past. This new computer was simply given the task to use its neural network to teach itself the game and it reached the point where it was able to beat the previous Ai 100 to 0 every time and in seconds. This is exponential growth at its purest.
The year after Google's AI beat the Go player, they created a new computer that was not schooled in the rules of Go and fed all the previous games of the past. This new computer was simply given the task to use its neural network to teach itself the game and it reached the point where it was able to beat the previous Ai 100 to 0 every time and in seconds. This is exponential growth at its purest.
Comments