To present AI as optimistic or dystopian? “That was the biggest argument”

advertisement

AI 2041: Ten Visions for Our Future is an unusual book. Each chapter consists of a short story, penned by science fiction writer Chen Qiufan, and a related analysis piece from Kai-Fu Lee, CEO of Sinovation Ventures and author of the nonfiction bestseller AI Superpowers. Chen, who also is founder of Thema Mundi, a content development studio, spoke with Fast Company on the eve of the release of AI 2041 about his collaboration with Lee, his own experiences with artificial intelligence, and what machine learning will mean for artists and writers. This interview was edited for length and clarity.

advertisement

advertisement

Fast Company: How did this project come about?

Chen Qiufan: I used to work for Google [from] 2008 to 2013, overlapping with Kai-Fu. Two years ago, he reached out to me. He had this brilliant idea of writing a book, blending the genre of science fiction with technical analysis, which sounded fantastic to me because I had the same idea years ago.

Before you started collaborating, did you or Kai-Fu Lee have a model in mind? Were there other works that you thought would be good models for what you wanted to do?

advertisement

No, this is something brand-new, because it is a combination of science fiction and nonfiction. So we had to build everything from scratch. It took us quite a while to figure out how to do it properly. The first half of the year was quite painful in my opinion, because I’m used to writing by myself, like playing solo. So it’s totally different, like playing in a band. We had to [go] back and forth to, to nail down everything. That took a while, but I’m quite happy and excited.

Can you talk a little bit more about that process?

Chen Qiufan [Photo: Yilan Deng]

Not long after we started writing the book the pandemic [emerged], so we had to work remotely. I am based in Shanghai and he usually works in Beijing or Taipei. We communicated by Zoom and conference calls. And before that, we had a lot of conversations with scientists and entrepreneurs, scholars, researchers on artificial intelligence. Then we had [to figure out] the right order to put all this different technology, where to go deeper and how to package different technology points and into [each] story. When we figured out the appropriate way [to tell] the story, and I started to write the draft and then Dr. Lee gave me some feedback and after several rounds of discussion and we finalized the story and handed it to the translator and the editor.

advertisement

Kai-Fu Lee has a generally optimistic view about artificial intelligence. Yet some of your stories highlight some of the darker parts of AI. How did you reconcile your his optimism with your artistic desire to paint a more rounded picture?

That was the biggest argument at the very beginning because in science fiction we tell a lot of stories about AI in the dystopian way, like The Terminator, 2001: A Space Odyssey, or Ex Machina. But in this book, Dr. Lee and I agreed on certain levels that we would try to build up this positive and bright future of AI and how it could empower the individual and societies. The problem is that if everything has a happy ending there’s no dramatic conflict, there’s no storytelling. So we had to figure out how to combine this kind of positive [viewpoint] together with some character development. And also we put [AI] into the context of a specific culture, like in India, Nigeria, China, Japan, America, Australia, and the Middle East. That kind of combination—bringing out some nuance and authenticity of how the technology might interact with the locality of the people and the culture—that’s how we figured out where the drama came from.

One of your short stories, “State of Trance,” included passages that were generated by AI. Where did the idea come from?

advertisement

Kai Fu Lee [Photo: Kai Fu Lee]

Back in 2017 I was writing a collection called The Algorithms for Life, which is six short stories about the relationship between humans and AI. I was thinking that I could leverage AI as a tool in this book. So I asked my ex-Google colleague, Mr. Wang Yonggang, [head of Sinovation Ventures AI Institute]. He works with Dr. Lee, and he is also a science fiction fan. He said, “okay, I’ll help you to build this algorithm.” Back then it was just LSTM (long short-term memory) and CNN (convolutional neural network), very simple stuff. We built up this model and fed it with all my writing materials to somehow mimic my style of writing, but I have to say, it made no sense at all. Last year we upgraded to GPT-2 [an AI tool created by OpenAI[. It’s a more advanced algorithm with much more powerful computation power. We fed it with tons of gigabytes of data [from] across the internet so it became smarter and it would write smoother and even generate something outside of your expectations. Right now I’m using it a lot of the time when I need to come up with some ideas that are not from me, but are from me on certain levels. So I use it as a tool, as a assistant to help me to come up with some fresh ideas.

What would you say to other artists, novelists, musicians, or visual artists who are inherently skeptical of AI, and how would you respond to people who say, “well, I’m afraid that AI will take away artistic or creative jobs”?

AI 2041 actually taps into a little bit of this, how artists might [evolve] in the future with the help of AI. We have technology such as natural language processing and AR (augmented reality), VR (virtual reality), and XR (extended reality). We can build super-realistic avatars, which can really recreate a real actor or actress on some level. In the future, we might use AI to assist or even replace real writers, musicians, and actors—average performers. If you reach on the top of the pyramid, [you’re] irreplaceable because of your imagination, your experiences in life, your perception of the world. Your value system isn’t replaceable by the machine. So I think the most important thing for all these artists today is really to find your own unique voice. Don’t copy anyone else. You have to be yourself. That’s the only way you can survive in the future with all this super-powerful AI robots. How we’re going to survive is with our imaginations.

advertisement

Were you already writing science fiction while you were at Google?

I was a science fiction fan since I was a kid. I watched Star Wars and Star Trek when I was around 7, 8, or 9. So that’s the age I start to read and write science fiction. I was an amateur when I worked at Google, so I used my 20% time on my writing.

Would you consider this kind of collaboration again?

advertisement

To be honest this is the most difficult book I’ve ever written, so I’d have to think about it very carefully, because not all collaborators are like Dr. Lee, who has such a great experience and expertise on a specific domain. He is also a creative person as well. Don’t be [fooled] by his super-professional businessman outlook! He’s got a lot of great ideas.

What’s an example of one of his ideas that you incorporated into your part of the book?

AI 2041: Ten Visions For Our Future by Chen Qiufan

One of my favorites stories is called “Contactless Love.” It [takes place] in Shanghai and it’s about the pandemic. It’s about the girl who has PTSD long after the pandemic. In 2041 she is still so afraid of stepping out of her apartment to find her lover who flew across the world to São Paulo. At the beginning [of writing] I thought, “okay, let’s just have a happy ending, have the girl meet the boy.” Dr. Kai-Fu Lee said, “why don’t we make it a game that was set up by the boy to guide this girl out of her traumatized status, like a gamification of therapy to guide this girl toward her lover.” I thought that was a totally brilliant idea, and I embedded it in my storytelling. And I think that it’s one of the most beautiful stories I’ve ever written.

advertisement