Technology futurists have been predicting for decades that AI (Artificial Intelligence) would eventually overhaul how we do our jobs, how we travel, and even how we engage in sex. But I’d rather not delve into my brother-in-law’s fascination with his AI robot girlfriend, the Monica XL-400. I still don’t understand their relationship, to be honest.
In recent months tremendous advances in AI have taken place, particularly with the introduction of something called AI chatbots. The one getting the most press coverage is ChatGPT from Open AI. So what, exactly, is ChatGPT? The GPT is short for generative pre-trained transformer. I am sure you found that explanation as helpful as I did. But as best as I can explain it, it is simply an AI-powered chatbot. We’ve all seen chatbots before. You know, those annoying online chat programs where you type in a request like, “Can I talk to someone in customer service?” and the chatbot replies, “Hello, I’m Brad. How can I help you?” And then you reply, “I just need to talk to a LIVE person,” and it replies, “Hello, I’m Brad. How can I help you?”
The technology of these AI chatbots is actually quite impressive. Just log onto one of these websites, type in any conceivable topic you’d like information about, and voilà! Within seconds, this AI program will compose a written response in surprising detail, crafted in such a way that it is almost impossible to tell that it was not written by a person.
ChatGPT, Chatfuel, Drift, MobileMonkey, and the slightly creepily named chatbot program, It’s Alive, are all exploding in popularity because they’re free and anyone can use them. If you know how to type and spell, you’re good to go – which is why my cat Zippy will most likely never use any of these programs, because the last time I checked, he still lacked opposable thumbs or any comprehension of what his name was.
Now, for the first time, there are utilities that will answer virtually any question in enormous detail. Sure, Amazon’s Alexa and Apple’s Siri do sort of the same thing, but the latest evolution of AI apps has taken everything to an entirely new stratosphere, as these chatbots can now use artificial intelligence tools to produce detailed text, images, sounds, and even videos that look and feel like they were created by humans.
Popular applications of this breakthrough technology include using it to write press releases and legal briefs, and helping to ensure that thousands of hardworking tech support and customer service employees all over the world will lose their jobs forever.
But probably the source of greatest controversy is the use of these programs to write high school and college term papers. You’re probably asking yourself, can a chatbot really craft a term paper that could fool a teacher? You be the judge. We asked a high school senior and ChatGPT the same question: “Who was George Washington?” Below are their responses. See if you can guess which response was written by Jordan Carruthers, a senior at Garfield High School, and which was provided by ChatGPT.
ChatGPT or Jordan Carruthers?
George Washington (1732-1799) was an American political leader, military general, and Founding Father of the United States. He played a crucial role in the American Revolution, leading the Continental Army to victory over the British and securing American independence. After the war, Washington was a key figure in the drafting of the U.S. Constitution and became the first President of the United States in 1789. He is often referred to as the “Father of His Country” for his leadership in both the military and political spheres, and his commitment to establishing a strong, stable government for the newly-formed United States. Washington’s legacy as a leader and statesman has been widely celebrated throughout American history.
ChatGPT or Jordan Carruthers?
George Washington chopped down a maple tree and had wooden teeth. He was very tall and wore a white wig. He stood in the front of a rowboat crossing a river in wintertime. That’s very dangerous, as he could have fallen overboard and drowned. He’s the guy on the one dollar bill. Did I mention he was tall? We get the day off every year on his birthday, so he must have been a fairly important dude.
Who wrote which version? It’s hard to tell, I know. Believe it or not, the first response was supplied by ChatGPT. But in fairness, Jordan Carruthers is a lazy idiot who’s probably not going to graduate on time. Of course, experts are concerned students may cheat and use ChatGPT to complete their term papers for them. This is a serious issue, and I, for one, am furious… that this technology was not available for me to use when I was in college back in the 1970s. So unfair.
Besides the obvious ethical concerns of the potential for widespread automated plagiarism, there are other significant challenges yet to be worked out. This technology has been shown, on occasion, to provide wildly inaccurate answers to questions – which could be problematic if you’re, say, a heart surgeon and you just asked the AI chatbot what to do next in a delicate coronary angioplasty and stent implantation, and the chatbot directs you to surgically attach the patient’s left leg to his right shoulder blade. Oops.
Another disturbing challenge in the rapid deployment of AI chatbots is that they can quickly become a**holes. Let me explain. AI chatbots are essentially highly sophisticated robot parrots. They quickly learn to assimilate knowledge – and opinions – based on the input they receive from the humans interacting with them. Before long, they start repeating the sentiments of their users.
So, imagine the tech team’s surprise when Microsoft launched its cutting-edge AI chatbot program called Tay. Twitter users conversing with Tay started tweeting the bot with a barrage of vitriol, including misogynistic and racist comments laced with offensive expletives. Within hours, Tay’s R-rated commentary started to make Donald Trump look like Mother Teresa. Guess it’s back to the drawing board, Microsoft.
I can appreciate that there are still a few bugs to work out before ChatGPT and the other AI chatbots become widespread in their adoption. But I for one am excited about the future potential. In fact, I am so impressed with these chatbots that I decided to use ChatGPT to compose this week’s entire column. From now on, instead of wasting seven exhausting hours working on my next column, I’ll just have an AI chatbot compose it. You can find me at the gym on the elliptical.
Next week’s topic: “The history of baseball.” I can’t wait to see what ChatGPT comes up with. I’m sure it will be compelling reading. I just hope it won’t be way better written than my usual columns.
That’s the view from the bleachers. I might be off base. If so, blame it on ChatGPT.
PS: If you enjoyed this week’s post, let me know by posting a comment, giving it a Like or sharing this post on Facebook.
Subscribe to my new View from the Bleachers YouTube Channel and request notifications to see my latest videos. And check out my new book, THE SECRET TO SUCCESS AND HAPPINESS (is Something I Have Never Figured Out. I’ Open to Suggestions).
I’m glad you brought up Siri. Is it my imagination, or does she sound annoyed when i ignore her direction to “turn left at the next light”?
Funny column and very close to actuality.
So is this from the science channel or the horror channel. Either way you make is sound like it might be an actual news clip. Good work, but I will keep my smart fridge and smart toaster locked up in case they cooperate. Who knows what will happen. Cold toast or toasted ice cream.
I’m curious Tim if your Chat GTP can be programed for sarcasm ? No one does it better than you