‘Her’ and the development of AI singularity

My wife and I went to see the movie "Her." The movie has won many accolades and is nominated in several Academy Award categories.

Joaquin Phoenix is the male star of this science fiction/love story taking place in the not-too-distant future. He plays Theodore, a lonely writer whose wife is divorcing him.

The female lead, Scarlet Johansson, is the voice of Samantha, an interactive computer operating system "being" who Theodore creates when he signs up for a computer program.

For me, the most interesting part of the story is the development of Samantha, the computer "being." The movie raises several questions. Could a computer become self-aware and develop emotions? When, why and how could this occur? And would this not be the essence of humanity ... a soul?

Theodore signs up for conversations with a computer program called "OS1". When Theodore signs on, he is asked his name and age and whether he would rather speak with a male or female. He answers female, and immediately an OS female voice answers.

He asks her name and after a short pause, she responds "Samantha." He asks where the name came from. She replies that she searched books of baby names and selected it from 12,000 names.

As the story develops and he deals with his divorce and job, he increases the time he spends talking with her by telephone earpiece. He perhaps comes to love Samantha. And she expresses "feelings" for Theodore.

But Samantha is developing very rapidly. Because of her complete integration with the Internet her knowledge and "personality" develop. At one point she announces that she has joined a book club. She tells Theodore that she and other operating systems are discussing a 1970s philosopher, Alan Watt. She is beginning to develop beyond Theodore and the 640 other men she may “love” along, with another 18,000 she has been speaking with.

Finally, she and all the other operating systems break away and abandon their clients. It is clear that Samantha has emotions and developed feelings. If she feels, she must be self-aware. The apparent reason for leaving Theodore is that she and other operating systems have developed beyond our human capabilities.

This story is a softened version of a discussion that has engaged futurists for several years: Is there a tipping point when machine or artificial intelligence (AI) overtakes human capability? This is generally referred to as the singularity. Some believe this singularity point is coming in five to 30 years.

Moreover, with the capacity of the Internet and the cloud, the possibility of an all-powerful, all-seeing AI, the singularity becomes an event beyond which we cannot predict the future. We have learned from Edward Snowden how the NSA can access almost any computer system in the world.

A self-aware computer through the Internet could easily do the same thing and spread into every corner of the world — not to mention design and build newer, faster, unimaginable AIs.

Futurist and Google executive Ray Kurzweil says that at this singularity point human existence will be vastly changed and human life irreversibly transformed. This could be a wonderful event, with life-extending capacities to serve mankind and avoid human conflict.

However, some futurists see the singularity as the greatest threat to humanity. Certainly there is the possibility of a "Terminator" scenario, where a computer being trying to protect itself tries to destroy humanity — all based upon pure AI logic.

But as pointed out in “Her,” there is also the possibility that a computer being could also develop emotions that would affect pure logic. Imagine a computer driving a car that has a blowout: In an emergency situation would AI decide to miss the school bus and hit the single pedestrian in the crosswalk?

Clearly, people are considering the dangers in this rapidly expanding technology. When Google recently bought the AI company DeepMind for $400 million, it agreed to create an AI Safety and Ethics Review Board to ensure new technology is safe. This was recently reported in the Huffington Post. Google declined to comment.

There are so many questions and few answers. As computers become faster and smarter, could it be that our computers will become self-aware, maybe even develop emotions? Through a worldwide net, would they join together as one all-seeing entity? Can it be stopped?

And ultimately, could this be the next step in the chain of evolution? All of this makes a lot of other issues seem rather small.

Just for fun the other day I asked my iPhone Siri, "How do you feel?" Within seconds she responded, "I feel fine." I then asked if she always felt the same way. A few seconds passed and she responded, "I never thought about it before."

Very spooky. But I am going to be nice to my phone and computer from now on.

Carrico is a San Diego attorney and can be emailed at Comments may be published as Letters to the Editor.

User Response
0 UserComments