Human-like Computers By 2029?

Human-like Computers By 2029?

Could Ray Kurzweil be right?

July 29, 2019, Bruno Jacobsen

The Famous Prediction about the Future of Artificial Intelligence


Ray Kurzweil is an American inventor, futurist and national bestselling author.

 

He has written seven books and has been named as one of the greatest inventors and most fascinating entrepreneurs in the country.


He is also famous for making predictions - some bolder than others. And, according to him, on mostly getting it right. 

In 2014, he predicted that computers would be at our level by 2029. At least in terms of emotional intelligence. That is, they will be romantic, funny, loving, and so on.

He further added that they would be customizable to each user. Depending on the personality and preferences of the user, artificial intelligence will respond accordingly.

Eventually, we can even build a real human-computer relationship.

But before that, the computers would have to pass the famous Turing test.

The test is simple: have a person in front of a computer and initiate a chat with it.

If that person, asking any questions they wish, cannot tell whether the responses are generated by another person or by a computer, the computer has passed the test.

In other words, artificial intelligence has become indistinguishable from a human. 

 

Will we have human-like computers by 2029?
 

Some people claim it already has. Last year, Alphabet's Chairman John Hennessy claimed that Duplex, an artificial intelligence voice technology, released by Google, passed the Turing test.



In the video, Duplex calls a hair salon to book an appointment. It sounds human; it reacts as a human would, and responds flawlessly, even when the desired appointment time is not available.

For now, Duplex would be mainly used for booking appointments and making reservations. And it passes the Turing test only in that domain.

It does, nevertheless, suggest that other domains, too, are up for grabs.

Mitchell Kapor, an American entrepreneur and founder of Lotus software, thinks we're still a bit far away from human-like intelligence - or at least emotional intelligence. He even placed a bet against Ray Kurzweil, which you can follow here: http://longbets.org/1/.

According to him, in a real Turing test, without restriction on the subject matter, everything encompassed by human experience would be fair game. 

It means a computer would have been able to respond to our concept of art, history, science, social relationships, our shared understanding of a soft summer breeze on a warm day; and linkages between the different realms.

For a machine to achieve this level of awareness poses a much grander challenge than getting an AI to book appointments in a human-like way, although the latter appears to be a step in the right direction.

One cannot conclusively say whether we will have human-like computers by 2029. Or if we ever will. With a decade left, the sensitive and playful AI "Samantha" from the movie Her still seems somewhat far-fetched. But, with technology advancing exponentially, one can never discount it either. 

As far as we know, passing the Turing test by 2029 seems possible, given what the test entails. Whether the Turing test is a good indicator of computers having achieved artificial general intelligence, that's a different story. Nevertheless, in less than a decade, we'll find out whether we have human-like machines or not. It's certainly a phenomenon that both individuals and organizations want to keep in mind.
 


Futures Platform is the #1 collaboration foresight tool that enables you to design the future, keep tabs on the latest developments and trends, and keep your organization future-oriented. 

Want a sneak peek at the Futures Platform tool? Free Registration HERE.

Leave your comment below: