Are you ready for a machine that's smarter than man?

How is that possible? The author believes that intelligence machines could outstrip human computer scientists in improving their own capabilities.

The analogy he brings to bear is frightening: The fate of the gorillas now depends more on humans than on the actions of gorillas themselves. It follows that the future fate of humanity could depend on the actions of machine superintelligence.

The outcome? Quite a few prominent thinkers have already predicted an existential catastrophe.

Barely a year had passed since the book’s publication when a team of researchers from the Massachusetts Institute of Technology (MIT) and the Computer Science Artificial Intelligence Laboratory (SCAIL) claimed on October 19 that they had developed a unique computer system based on artificial intelligence that can “outperform even the smartest humans on Earth”.

Now, that is frightening, to say the least. Researchers have taught machines to “self-learn”, placing us close to the point of no return.

In May this year, British cosmologist and Astronomer Royal Martin Reese warned that an explosion in artificial intelligence has sent us hurtling towards a post-human future.

He wrote: “We’re witnessing a momentous speed-up in artificial intelligence (AI) – in the power of machines to learn, communicate and interact with us. Computers don’t learn like we do: They use ‘brute force’ methods. They learn to translate from foreign languages by reading multilingual versions of, for example, millions of pages of EU documents (they never get bored). They learn to recognise dogs, cats and human faces by crunching through million of images – not the way a baby learns.”

The MIT-CSAIL research team’s AI breakthrough was based exactly on that: Predicting patterns buried in unfamiliar data sets.

The team’s Data Science Machine (DSM) sifts for patterns in data sets, such as promotional-sale dates and weekly profits. Before the DSM, it had always been assumed that, while computers could perform many tasks faster than humans, machines would still require our input to choose what to find in a huge data set.

In other words, machines need humans in order to find meaning in patterns.

But the new computer system developed by MIT-CSAIL has now automated that process too. The scientists believe they have probably cracked the code.

The Data Science Machine recently outperformed 615 human teams in three competitions of pattern-finding ability. The human teams spent months developing algorithms that would predict patterns for the task outlined in the competition. The machine, it turned out, was able to compute the prediction in 12 hours at the most – and sometimes as little as two hours.

This breakthrough could complement human intelligence in a number of tasks. The researchers cited the example of analysing statistics to predict the likelihood of students dropping out of an online course.

On this evidence, however, the robots aren’t taking over the world just yet. In fact, this latest evolution in AI has the potential to solve many more real-world problems. There are endless sets of data out there waiting to be analysed by the New Machine.

My recent encounter with a robot called Nao in Japan was friendly and instructive. “Tell me what you want me to do, give me enough data and I will do it for you,” the robot seemed to be telling me. 

Then, when it fell flat on the floor after his handler inadvertently pushed it aside, the robot declared: “Oh, I slipped. I need to get up.” That convinced me that Nao was no mere “obedient machine” any more.

Martin Rees writes of AI: “There is disagreement about the route towards human-level intelligence. Some think we should emulate nature and reverse-engineer the human brain. Others say that’s a misguided approach – like designing a flying machine by copying how birds flap their wings. But it is clear that once a threshold is crossed, there will be an intelligence explosion.”

That “intelligence explosion” will come about, he declares, because “electronics is a million times faster than the transmission of signals in the brain; and because computers can network and exchange information much faster than we can by speaking.”

The next time I meet Nao, I’ll tell him: “We need to have a serious conversation about our future relationship.”