I specialized in artificial intelligence (AI) and simulation in grad school. Even in those nascent days of AI, preparing for the future with a foot in the present seemed like a sensible ticket to a paycheck. The class on semantics was the most intriguing. How can a machine interpret what we mean and not what we say? Back then, we were connecting the dots with linear discovery.
In the last four decades, the tech industry has traveled on an exponential arc to weave a complex fabric of vision, speech, touch, learning and intellect to shake the world with the incredible possibility of AI and the accompanying doomsday scenarios.
We have quickly gone from ho-hum on AI to dire warnings about it.
Noted researchers and corporate stalwarts - including the CEO of OpenAI - are warning of human extinction in the scale of pandemics, nuclear wars, and climate catastrophe if AI goes unchecked. Some of us are sitting up and listening, others are rolling their eyes.
Question is: what can AI actually do today?
AI can see. We know this from self-driving cars and facial recognition apps. It can view, react, learn and adjust.
AI can hear. Alexa and Google Assistant anyone? AI can hear a sound and know what it is. It can discern words and intent even over incoherent background noise.
AI can smell. AI-driven digital olfaction is mimicking the sense of smell by capturing odors using biosensors and then modeling the signals for managing diseases such as Parkinson's and cancer.
AI can touch. Think robots on rails picking up objects. Now picture robots deciding whether a fruit is ripe or not. AI can digitize the sense of touch, recognize transparent and opaque objects and combine tactile and visual signals to emulate human reactions.
AI can taste. Given consumer data on reactions to different foods, it can predict and optimize flavors for a particular market. For instance, winemakers are modeling individual preferences and wine profiles with high accuracy using AI.
If you are physically challenged and AI can replenish your senses, imagine the possibilities. If you are in national security, disaster management, human resources or large scale manufacturing, think of how you can scale with precision using AI. If you want to rest while you drive, rejoice at the mainstreaming of fully self-driving energy efficient transport. If you are a doctor, reflect on the boundaries you can break with rapid detection with AI.
It's a no-brainer that we need to go where AI is headed. So, why is everyone hitting the panic button?
Because AI is also "thinking".
AI is creating content that so far only humans have delivered. It is dynamically responding to our questions with an incredible grip on semantics. It is writing marketing content, high school and college essays, solving complex math problems, creating poems and visual arts, composing music and debating us. It is researching, positing and rebutting in real-time. ChatGPT has dramatically lifted the curtain on the evolution of learning algorithms and generative AI that now thinks.
AI is also reading our minds.
Now you can raise the alarm!
AI is reading brain waves and translating them into words while we are thinking. Meaning I can think of the word "crap" and AI can speak it aloud. AI could have harnessed Stephen Hawking's brilliant brain waves and turned them into a stream of consciousness without the use of his thumb and eye blink!
The good-for-mankind applications, the nail-biting crime solutions, the cost-effective economics are certainly not causing panic.
When (not if) this thinking AI falls into the wrong hands and a nuclear war is initiated because of an evildoer's thoughts, imagine the catastrophe. Given that we can barely deal with human biases created by politicians and the media, consider AI-fueled bias that you cannot rebut. What about our privacy when machines know what we are thinking or create unauthorized content on our behalf? What happens when an AI-driven sell-off creates market volatility and wipes off our hard earned assets? How do you parse deepfakes where your face is put on another body and you are digitally made to do the unthinkable and then law enforcement fails to differentiate between our real and AI-generated personas?
It could take a while to get to these edge scenarios or we may never get there. But then again, did we think Alexa and Siri would be listening in on us 20 years ago?! Chaos is always closer than we think and technology never stagnates or regresses.
To mange the harmful uses of AI, experts are calling for immediate controls because unchecked exhilaration will set us on the road to anarchy and panic will keep us from making progress. Now is the time for policies and collaborative actions like the ones that have prevented nuclear disasters after Hiroshima and Nagasaki.
As for our paychecks, our intersection with AI is inevitable regardless of what we study and what we do in the workplace and how we operate at home. Whether we are young or old, creators, producers or consumers, sooner or later our story with AI is going to take shape. If we are not ready, we will get in queue for a job loss or fail to cope with the pace of change.
With proper constraints in place, the hope of leveraging AI in the right capacity and thriving in the next chapter of human history rises exponentially. Clearly, it is time to pivot and lead AI in the right direction. Godspeed to the ones who will manage this change and may the rest of us be able support in the best ways we can.