We Need to Stop AI Before it Replaces Tech CEOs Like Me 

Let me start by saying nobody is a bigger believer in technology and innovation than I am. But just like you, average person, I sometimes sit in my aqua-bubble fallout shelter at the bottom of the ocean and wonder if we've gone too far.

After watching the first 12 minutes of I Robot, I invested heavily in AI technology. It was thrilling stuff: seeing the first frustratingly circular customer service texts break someone's will, watching a program scrape millennia worth of art images from the internet and spit out a digital painting that looked like it belonged on a van in 1977. When we created the first AI musician, making her look like a hot teen was my idea. I was there when the first chatbot wrote a new speech in the voice of Abraham Lincoln, but, like, gay. Heady stuff imagining what we could create once we jettison the human artists, musicians and writers. But then I pondered what the world might lose if I were replaced by AI.

When factory robotics first took off, everyone worried about workers losing jobs. But with AI, even CEOs and owners are in danger. An assembly line of robots can make a Ford, but now that we've fed AI from the darkest corners of 8chan, couldn't it also match Henry Ford's virulent antisemitism? Would an artificially intelligent entity deploy police and hired thugs to beat and fire upon striking auto workers as Ford did? We can hope but we can't be sure.

AI cannot be trusted to make the kinds of calls a tech CEO handles. Black turtleneck? Ribbed black turtleneck? How many timed bathroom breaks for warehouse employees?

There's an intellectual freedom a tech CEO with little to no oversight has that can't be replicated. AI is capable of reinforcement learning, whereas, if you look at my social media feed and my personal and professional track records, I clearly am not. In my genius neural network, new ideas are completely divorced from past experience or expert input. And with zero capacity for self-reflection, I'm equally surprised every time a business partner, girlfriend, advertiser or offspring unit stops speaking to me. And that makes me dangerous. But in a cool way, not like self-driving cars.

When asked a question outside their training, AI programs sometimes give false answers with a tone of complete confidence, similar to my takes on world politics or women. Except with AI, these are called "hallucinations." Please. I have actual stock market-rocking hallucinations, like when I was on diet of 100 grams of hydroponic lichens a day and instructed my rocket guys to pivot to drilling into the Earth's core to find my underworld doppelganger.

AI might be able to analyze the entire global market in seconds and fully comprehend what my coders and engineers are doing at all times, but that's not leadership or camaraderie; that doesn't build a workplace culture. It takes a man like me to look another man in the eyes and ask if I've fired him yet, or to ask a woman by the copier what her IQ is and if she'll gestate my baby. An AI-driven robot will never do more than mimic the lifelike facial expressions a professional likeability coach has helped me master. Look at my incredulous eyebrows lifting a full centimeter at the very suggestion.

Sure, generative AI can recognize and mimic certain patterns, like layoffs and factory closures after a new crop of hair plugs go in, but it's not the same. Because when it comes time to hastily throw together a prestige biopic, you won't be able to attribute an AI tech mogul's tantrums to sociopathy or the lonely little boy inside. It'll just be ones and zeroes that cranked out all that chaos. That's not the future I've been getting transfusions of young people's blood for, to say nothing of the spare organs I've been growing in an underground lab.

I'm still deploying facial recognition software to read the micro-expressions of warehouse employees to see if they are having negative thoughts about me. But we've got to draw a line before AI takes my spot. Do you really want a soulless AI entity mining your personal data and enabling genocide on the other side of the planet when you could have a soulless middle aged human man handle that? You don't have to answer — the software can tell by your face.

Jennifer Fumiko Cahill (she/her) is the arts and features editor at the Journal. Reach her at (707) 442-1400, extension 320, or [email protected]. Follow her on Instagram @JFumikoCahill and on Mastodon @jenniferfumikocahill.