facebook twitter instagram linkedin google youtube vimeo tumblr yelp rss email podcast phone blog search brokercheck brokercheck Play Pause
My New Assistant, ChatGPT Thumbnail

My New Assistant, ChatGPT

Photo: Courtesy of Owen Beard (Unsplash)

Artificial Intelligence is really advancing fast

By Jon Aldrich & ChatGPT

Okay, It is Monday and I am feeling a little lazy today in writing my blog, so I am going to have ChatGPT, the free, online chatbot help me write this article. If you have not played around with this amazing Artificial Intelligence (AI) tool yet, you might want to give it a try. It is astonishing, but yet a little bit scary as to how far AI is advancing. So, ChatGPT and I sat down and wrote this together. I started out by asking it write a paragraph about what it is:

Okay, that was kind of cool, I can type in a question and get a detailed answer, much better than searching Google and trying to separate the wheat from the chaff. But how does my new associate, lets just call him, Chat, come up with this information? Apparently, Chat isn’t just going to Google and compiling search results, it is trained by massive amounts of data that researchers have gathered from the internet and other sources the last several years. The answer it comes up with is derived from the vast knowledge it has accumulated and continues to accumulate, and well, “learn” as time goes by.

I also asked my new friend, Chat, if he could write an article about what exactly it is and how it might be used for and within seconds I had several paragraphs describing in detail what it is all about. I now see why kids in school are writing term papers using this new tool!

Okay, I can see this is going to be fun and that sure saved me a bunch of time researching and typing up. Heck, I think its grammar is pretty good too. The more you interact with it, the more it almost feels like a conversation. You can see where this is going to be great for large companies online help lines and how much potential there could be. (We will be cussing at Chat, rather than some poor person in India!)

Next, I wanted to see if Chat, had a sense of humor and he didn’t disappoint and he has an inventory of jokes:

Okay, then my immature side kicked in:

So, my pal, Chat, does have some boundaries and keeps things clean. Nice. Next, I wanted to see if Chat is going to replace me soon:

Okay, it looks like he still needs me as it is a generic answer and he is recommending to consult with an advisor, so maybe I can be useful for a few more years. The more I played around with it, the more I figured out, Chat was good at answering specific questions, as long as they were not too new since the data only goes through 2021. So when I asked it, “Why is high inflation so bad?” it gave me 5 great reasons and explanations of the effects of high inflation which could easily be cut and pasted into a term paper or a blog. However, when you ask it something that may not have such an objective answer it does leave a little bit to be desired:

So, it was not the worst answer, but it was a little more generic than what I was hoping for. But, you can provide feedback to help it “learn” going forward. Thus, it may ask when you provide feedback, What would the ideal answer have been? You can provide some input that may be considered in the future if the question is asked again.

I had a lot of fun spending some time asking tons of questions trying to find its strengths and limitations. I think it can replace Google in many respects for something like a quick recipe, (it was great when I asked for a cookie recipe or the best way to prepare pork chops) or finding out actual facts. It still has a long way to go when you ask it more nuanced questions that are more subjective as it is trained to provide neutral responses based on the information it has been trained on and patterns in the data.

All in all though, you can see where AI is headed and although it still has a long way to go, advances continue to come quickly, so what we are seeing with my pal, Chat now is probably just the tip of the iceberg and when future iterations of it arrive, it is going to amaze you at what it can do. Hopefully, the Terminator scenario in which AI tries to extinguish humans is a long way off, but you do wonder, can it ever evolve to that someday?