A popular movie on Artificial Intelligence, “I, Robot”, led to a discussion on the sentience of robots and how it will affect human life. Or the famous “Terminator” series, in which the mighty AI, Skynet controls the lives of the humans. While these both were dystopian fiction and an introduction to what Artificial Intelligence is for the laymen, it is far from reality on the current state of AI.
A recent uproar in media claiming that “Facebook shuts down AI which created its own language” being misleading, reflects to me the large gap between the research industry and the general public who grew up with movies like Terminator and I, Robot. To clarify things to one’s fearing Skynet invasion, I assure that Facebook shutting down this AI is a misleading media hype and without delving into nuances of the field, I will try to explain what actually happened. The researchers give a reward to AI (in form of points) for doing certain things correctly; in this case, it is speaking to another AI agent. Sometimes, the AI optimizes the outcome in wrong but effective ways; here it just used some shorthand version of English. The outcome required was not as expected, so Facebook changed the parameters of the experiment which the media claimed as the onset of Skynet invasion. Such failed experiments are very common in research and a post by CNBC clears the air around the controversy.
With clickbait-y articles such as the above, it is imperative that the non-researchers understand that we shouldn’t be worried about robotic overlords and that AI is actually going to improve human life. Imagine the world in which your car drives itself or where robots work alongside you to enhance the quality of your life. Or when your fridge does your inventory shopping. What will this future look like? How will it change our lives?
In the 2017’s TED conference, Tom Gruber, the co-creator of Siri, talked about wanting to make “humanistic AI” that augments and collaborates with us instead of competing with (or replacing) us.
He shares his vision for a future where AI helps us achieve superhuman performance in perception, creativity, and cognitive function. “We are in the middle of a renaissance in AI,” Gruber says. “Every time a machine gets smarter, we get smarter.” This is what a facet of AI looks like, something that enables humans to use the technology to improve their lives significantly.
Janet Baker, founder of Dragon Systems talked at a dinner at TED 2017 on how AI will transform our lives in future, “AI will continue to provide a set of tools to people that expand their horizons and enhance their ability to work and play” he says.
Also, a regulated and well equipped driverless car could potentially perform better than a human driver. Statistically speaking, the majority of road accidents are due to the least reliable technology in the car: the driver. While humans are error prone in certain judgments, the self-driving car system when well-regulated will significantly reduce the accidents occurring due to the error in human judgments. Chris Umron at TED 2015 took us through how driverless car sees the road.
Active research is going on at companies and universities around the world to improve the system of road scene understanding for autonomous cars.
Another field of AI research, on social robots, is aimed to advance the state-of-the-art socially intelligent robot partners that interact with humans to promote social and intellectual benefits, work alongside with humans as peers, learn from people as apprentices, and foster more engaging interaction between people. This has real-time applications in improving the childhood experience for kids, assisting in physical skill development and much more.
Customer-centric AI such as Amazon’s Go will revolutionize the shopping industry. All e-commerce retails like Amazon, Flipkart, Myntra etc. use AI to predict user’s preferences and show relevant advertisements.
While only a few branches of this huge industry has been introduced here, it is clear that the field of AI and its sub-fields will prove useful in enhancing and complementing the human abilities.
While there is a certain doubt in general public about how effective these systems are and whether we are programming ourselves into oblivion, TED brought up this question at a dinner hosted by Toyota on the future of AI and the excerpt of response can be read below.
“Powerful technologies will be used and abused. Sophisticated AI-based technology for pattern recognition can be used to recognize the words we speak, faces in crowds, cancer cells in images, or protective radar signal analysis. It can also enable the automated surveillance of vast quantities of audio and visual materials, and unprecedented profiling and tracking through the collection and convergence of personal data. We must be aware and take active roles in advancing our capabilities and protecting ourselves from harm––including the harm from escalating prejudices we foster by isolating ourselves from differing ideas (e.g., with polarized news feeds) and productive discourse about them.” — Janet Baker, founder – Dragon Systems