Is Conscious AI Achievable & How Soon Might We Expect It?

Artificial general intelligence (AGI) can be defined as artificial intelligence (AI) that matches or surpasses human intelligence. It is, in brief, the type of intelligence through which a machine is able to perform any intellectual task that a human being can. And, it is currently one of the main objectives of AI research.

The concepts of AGI and consciousness, however, lack definitions that satisfy everyone.

The type of artificial AI currently available is focused on specific tasks and is therefore referred to as “applied” or “narrow" AI because of the machines' limited intelligence.

The United Nations Educational, Scientific and Cultural Organization (UNESCO) recently published its quarterly journal which is entitled Artificial Intelligence: The promises and the threats.  It covers various sub-fields of artificial intelligence, including briefly touching upon conscious AI.  In the journal it is stated that AGI would be a machine that has consciousness and feelings, and would be capable of "providing solutions for any kind of problem – that is pure fiction, for now.”

Is conscious AI achievable?

At this point, it is possible to design robots that look like humans. Japan's Hiroshi Ishiguro, for instance, has created Geminoid robots that look just like humans. But, much more development of these types of robots needs to take place before they can be placed in a home as a housekeeper, for instance. Robots often fall down and some require human assistance with walking. Their movements are not fluid and they would also need to be able to recognize a myriad of objects typically found in the home.

To solve computational tasks quickly, advanced AI systems use a process known as deep learning. Networks of layered algorithms, that communicate with each other in order to solve increasingly more complex problems, are employed. At this time, however, neural networks are dependent on a human programmer who sets up the tasks and chooses the data the machine will learn from.

Consciousness for AI would mean that neural networks could make those choices themselves. Edith Elkind, a professor of computing science at the University of Oxford explains that, "machines will become conscious when they start to set their own goals and act according to these goals rather than do what they were programmed to do." Elkind points out that this differs from autonomy in which a fully autonomous car drives from A to B as programmed, but not based on its own decision-making processes.

"Once we can spell out in computational terms what the differences may be in humans between conscious and unconsciousness, coding that into computers may not be that hard," says Hakwan Lau, a UCLA neuroscientist.

Regarding the viability of conscious AI, Subhash Kak, Regents Professor of Electrical and Computer Engineering at Oklahoma State University has said that, “researchers are divided on whether these sorts of hyperaware machines will ever exist. There’s also debate about whether machines could or should be called “conscious” in the way we think of humans, and even some animals, as conscious. Some of the questions have to do with technology; others have to do with what consciousness actually is.”

How soon could we see conscious robots?

Some scientists are optimistic that it can be achieved within a few years, while others estimate it will probably take a few decades. Elkind, for example, believes that "we are very far from having conscious machines." Still others doubt it will happen at all.

Yann LeCun, Director of AI Research at Facebook and professor of computer science at NYU is also of the opinion that it will be some time before machines can learn the most basic things about the world in the manner humans do. “Like, yes, in particular areas machines have superhuman performance, but in terms of general intelligence we’re not even close to a rat.”

According to Luke Tang, general manager of AI startup accelerator, TechCode, once we are able to successfully move forward with “unsupervised learning algorithms,” it will be possible for machine intelligence to quickly surpass human intelligence.”

He estimates it's "probably only 30 to 50 years away.”

Pierre Barreau, CEO of Aiva Technologies, thinks that, "a lot of people don’t realize the complexity of our own brain, let alone creating an artificial one.” Barreau explains that, “in order to achieve AGI, there will need to be major breakthroughs not just in software, but also in Neuroscience and Hardware.” Continuing, he said: “We are starting to hit the ceiling of Moore’s law, with transistors being as small as they can physically get. New hardware platforms like quantum computing have not yet shown that they can beat performances of our usual hardware in all tasks.”

In order for AI to be considered truly intelligent, it is widely accepted that it must pass at least five tests, foremost of which is the Turing Test. The Turing Test involves a machine and a human both conversing with a second human being, who tries to determine which one is a machine.

Barreau has little doubt that we will see, in our lifetime, AI passing the Turing Test--as a human being.

In late 2017, cognitive scientists Stanislas Dehaene, Hakwan Lau and Sid Kouider theorized in a published review that consciousness is “resolutely computational” and subsequently possible in machines. The three neuroscientists are from the Collège de France, University of California and PSL Research University, respectively. They addressed the question of whether machines will ever be conscious:

“[But] the empirical evidence is compatible with the possibility that consciousness arises from nothing more than specific computations.”

In Japan, robots are already working as shop assistants, in personal care and in schools.

Genius concept or existential threat?

Esteemed physicist, Stephen Hawking who passed away in March, warned that the development of full artificial intelligence could lead to the end of humanity. Hawking told the BBC that it could signal the end of the human race.  He was very uneasy in regard to the consequences of creating something that could match or outpace humans. "It would take off on its own, and re-design itself at an ever increasing rate," he cautioned.

A (slightly) more optimistic viewpoint comes from Rollo Carpenter, creator of Cleverbot: "I believe we will remain in charge of the technology for a decently long time and the potential of it to solve many of the world problems will be realized."

China and Russia are apparently hoping to finally surpass the US in military dominance, by way of AI. Russian President Vladimir Putin has said that, “artificial intelligence is the future, not only for Russia, but for all humankind.” He added that, "whoever becomes the leader in this sphere will become the ruler of the world.”