Blade Runner 2049, the sequel to the classic Blade Runner, will be in theaters soon. I can’t wait, because the original was not only a technological tour de force when it was released in 1982, it also offered a critical perspective on the future of robots, artificial intelligence, and our possibly dystopian future.
Blade Runner explored our fascination with — and fear of — futuristic worlds where AI dominates and it is impossible to tell who’s who in the battle between good and evil. Because I’m in the AI business, I often ponder how realistic these cinematic worlds really are and what the media depictions get right and wrong.
Do you have an AI strategy — or are hoping to get one? Check out VB Summit on October 23-24 in Berkeley, a high-level, invite-only AI event for business leaders.
All AI is not created equal
The original movie’s dark vision is startling, but it does get a few things right. The “replicants” depicted in the film are an example of “strong AI,” or artificial general intelligence (AGI) that allows a machine to perform virtually any human task. These characteristics include the ability to make advanced, independent decisions and makes the replicants faster and stronger than humans. “Strong AI” also implies that robots could have super-human intelligence, though there is a debate as to whether this sort of AI-powered autonomy and intelligence constitute true consciousness.
The “humanizing” of AI is a big task. Applied AI, or narrow artificial intelligence (NAI), is able to solve specific problems and make a more limited set of decisions. NAI doesn’t seek to simulate full human cognitive abilities, but it can do things like recognize faces, play chess, and effectively carry on certain types of conversations. These narrow AIs — though highly developed — are currently unable to operate outside their focused area of expertise and often need structured methods for receiving data, as opposed to using traditional “human” senses for input.
Blending multiple AI technologies can enable a bot to solve specific problems and achieve conversational objectives. In the marketing world, natural language processing (NLP) and natural language understanding (NLU) can also be combined to interpret free-form text received from customers’ leads and contacts. Because people who send these messages typically don’t realize they’re conversing with an AI, they make no extra effort to limit their vocabulary or structure their messages. Finally, natural language generation (NLG) can be used to craft helpful and human-like messages that advance the conversation while achieving conversational objectives.
Fact or fiction?
When it comes to AIs, robots, and replicants, however, films get a number of things wrong — mostly to scare us. A brief examination of what’s actually plausible might allay some of those fears.
Despite all of our advances in AI technology, it is truly difficult to make an android indistinguishable from a human — in terms of both appearance and behavior. And even if androids achieved super intelligence, it doesn’t follow that they would necessarily experience the same range and types of human emotion and emotional intelligence that replicants in Blade Runner seem to display. It also seems unlikely that androids would be able to feel love for other androids or even interact socially with each other. Any sort of advanced emotional intelligence would likely require a separate and different programming approach, and such capabilities would only arrive in a later technological evolution.
Most importantly, any synthetic human/android would almost certainly include fail-safe programming, preventing it from directly or indirectly harming humans, as described in Isaac Asimov’s Three Laws of Robotics.
I see a future where AI can never replace humans. In fact, responsible AI and humans are stronger together than alone. AI assistants will continue to evolve quite rapidly, becoming more helpful in ways we can’t easily predict. Narrow AI will continue to advance and will probably be aggregated into meta-bots, like Amazon Echo, Apple’s Siri, and Google Home, or used in more general-purpose AI assistants that attempt to solve a growing set of problems. Predicting the future is hard, but movies like Blade Runner make it fun to try, ultimately helping us think about how AI can improve quality of life for everyone.
Alex Terry is the chief executive officer at Conversica, a company that creates conversational AI software for marketing and sales.