Raajeev Aggerwhil
By Raajeev Aggerwhil
When I tell people that I have an AI-start up, they roll their eyes making it clear they are thinking, “Not Another AI start-up.” Among those who know me as a stand-up comedian, think I am telling a new joke! When I mention that I studied AI at Stanford and Computer Science at Berkeley in the late 80’s, they are surprised. I know they must be thinking, “higher education at work.”
I am happy to see a resurgence of Artificial Intelligence (AI), in the form of Generative AI, a branch of AI capable of generating text, images and videos. Back in my undergraduate years at Berkeley, I was interested in not only cognition but semantics and intentionality behind the spoken words. That is one of the reasons behind the discordance in the first few years of my marriage! I was like a non-stop autocorrect; it must have caused a lot of irritation and frustration for my wife.
With the emergence of ChatGPT, and many similar Large Language Models (LLMs), my interest and passion reignited. I realize that I am as passionate about AI as my comedy. So, I embarked on an AI start-up with a couple of my Ph.D. friends. At an open mike, I was telling people, “I have good news and bad news. The bad news is that all of our jobs will be taken away. The good news is that it would be done by machines and not Indians through outsourcing. At least there would be a neutral accent.”
One reason behind my AI start-up journey, besides my passion and interest, was fear. If the jobs are going to be taken away, I would rather be on the other side… the side taking away the jobs. The last reason behind this new venture was that I wanted to conform and live up to the expectations set by my fellow Indians. When people are in distress about the loss of their jobs, they can pander to their biases and take comfort in blaming brown people in helping accelerate the job losses. It is like when you are going 75 mph on a highway and you see a car going 35 mph ahead of you, you immediately change lanes. And while passing the slow-moving car, you take a quick glance at the driver. If their gender, age or race conforms to the group you are most prejudiced against, you nod with relief and say to yourself, “Of course!”
For the AI start up, we are going to investors to pitch the idea to raise money. What is weird about doing this is that the skills I learnt as a stand-up comic are completely useless here. If I pitch investors and I manage to make them laugh, I am in deep trouble. “Ha ha! You want money for this idea. That’s funny. Tell us another one.”
Actually, I do find my comedy to be helpful sometimes, especially when I have to explain complex technical topics to a nontechnical audience. LLMs like ChatGPT sometimes have a problem called Hallucination. If the system does not know the answer or does not have any evidence, it will make it up. AI programmers have really struggled with this. I am very familiar with this problem. I have been married for over 25 years! I have had these kinds of discussions with my wife for years. The other day I was meeting an acquaintance for lunch. Later my wife asked me, “Did he ask about me?” I said, “Why would he ask about you?” She said, “Well, you always ask about other people’s wives.” Surprised I said, “When do I do that?” Like LLM’s, she didn’t have any evidence. So, she simply said, “All the time. Every time.”
That is why to combat hallucinations associated with LLM’s, the software developers design programs that have transparency and explanation built in. Unfortunately, in my personal life, I just need to learn to deal with minor hallucinations as a fact of life. If I insist too much on explanation or transparency, I’ll be in trouble!
When I pitch our AI offering to potential customers, they are concerned about the job loss. I tell them our software system is designed to enhance productivity and eliminate grunt work. If they are not convinced, I tell them to think back to 40 years ago when PCs with word processors like WordPerfect were coming in to replace typewriters. There were people who would resist technical change. Some were even afraid of technology. In fact, after the industrial revolution, in Great Britain, there were workers who would go about breaking the machines because they were afraid that those machines would take their jobs away. Those people were called Luddites. Several years ago, when I got my Tesla, I decided to get a customized plate, “LUDDITE.” I figured that name for an electric car would be a good expression of my humor.
The group of people in favor of efficiency, whether by machines, Information Technology and now AI have always faced resistance from people like Luddites. However, in the case of AI, a lot of people are afraid, and these are not blue-collar workers. A lot of knowledge workers, like accountants, financial analysts, even radiologists are afraid that their jobs might become obsolete. I think the concern is genuine and the best way to deal with it is to focus on the core skills of your job, get better at it by using technology like ChatGPT and other LLM’s as an aid to increase your efficiency. That way, you are indispensable.
I tried running some of my tried-and-true punch lines of my jokes I have had great success with to see if ChatGPT could do any better. It didn’t. So at least, I know one thing, my job as a comedian is not going to be replaced by AI anytime soon!
Raajeev Aggerwhil is a Los Angeles-based comedian. Follow him on Instagram @raajeevcomedy or @IndianComedyCouple. To see his videos on YouTube: youtube.com/c/raajeevcomedy.
