In terms of artificial intelligence – AI – the point of singularity is when artificial intelligence systems start building and improving upon artificial intelligence systems. The AI no longer needs human intervention and it can create more artificially intelligent machines. It’s like life creating life on its own. There may come a time when artificial intelligence systems may start deciding upon themselves when they need to develop more advanced machines and when they need to develop more machines.
This MIT link says that while people are worrying about AI robots taking their jobs like working in the factories, calculating insurance, attending to people at hotels, and even driving trucks, trains and aeroplanes, soon a time is approaching when software will be able to write more software and hence, AI robots will start taking jobs of less developed AI robots. Writing software is a highly complex ability and once artificial intelligence can create software it will be as good as humans becoming more intelligent and getting more knowledge to better themselves on their own. The link says that as more and more AI systems learn to develop more AI systems, the cost of building complex AI systems will drastically come down and they will be easier to implement across multiple industries:
If self-starting AI techniques become practical, they could increase the pace at which machine-learning software is implemented across the economy. Companies must currently pay a premium for machine-learning experts, who are in short supply.
Researchers at the Google Brain artificial intelligence research group already working on software that can design machine learning systems. So some of the software needed may be written by software itself. According to Jeff Dean who leads the Google Brain research group, “automated machine learning” is one of the most promising research avenues his team is exploring.
The MIT article further says:
The idea of creating software that learns to learn has been around for a while, but previous experiments didn’t produce results that rivaled what humans could come up with. “It’s exciting,” says Yoshua Bengio, a professor at the University of Montreal, who previously explored the idea in the 1990s.