Since late November 2022, ChatGPT has revolutionized the AI landscape. In this first episode of Season 4, I share the musings of my interactions with this magic tool.
Links to articles I mention in the episode:
Hello, peeps. It's been a while since you heard from me. Thank you for your generous support of my show. Thank you to all those who have reached out to me inquiring about my new season. Life events of the past few months have delayed the season. But here I am. With my first ever monologue for the first episode of season four. These are my musings on ChatGPT. What better subject to talk about right? Well, with the advent of ChatGPT, we have kind of catapulted abruptly into a very futuristic era, an era that at least I was not prepared for. Well, now that we are here, I want to share my thoughts on what I have gleaned from my interactions with this amazing artificial intelligence technology. So what is AI? What is artificial intelligence? In very simple terms, it is the ability of machines to think and learn like we humans do. In doing so they become intelligent. Now, how do machines learn and become intelligent? Artificial Intelligence is built with highly advanced software that learns from datasets using a set of algorithms. These algorithms are really a set of rules that use statistical models and computational techniques to analyze the data to identify patterns and gain insights to make predictions or decisions. Now, as this data set grows, over time, the software can continue to learn and improve its accuracy and effectiveness. Now if you're wondering where I got this succinct summary from, I will credit ChatGPT, because I brainstormed with ChatGPT to come up with this refined summary of how machine learning happens, and Artificial Intelligence systems work. Where can we find it? We have been interacting with AI for several years now. It's in speech recognition, image and video generation, natural language processing, and so on and so forth. Some daily examples include Siri, chatbots, Spotify, and even smart compose if you you're using Gmail. Like many others, I have had my share of prompts to ChatGPT that resulted in some amazing responses. I have used it to, like others have said, you know, to create outlines, to write summaries, and to brainstorm. Even the content that I'm presenting to you right now, a lot of that has been gleaned from my prompts to ChatGPT. Another interesting discovery that I had was, that I thought about was, hey, I speak several Indian languages. Would ChatGPT be able to respond to me if I use the Latin script to ask prompts in those languages. So for example, I posed a prompt to ChatGPT, using the Latin script, but in Kannada, and lo and behold, it responded to my prompt. I used several other languages and ChatGPT was right on point. This was very impressive. I think this technology would, I think, would probably help me learn a new language altogether. I also speak in languages that I'm familiar with, but I'm not fluent in so I thought, hey, there are these idioms that I've heard and that have, I have tried to, you know, use Google to get some information or some insights on the chat GPT was point on, it was accurate. And it gave me some insights into these idioms that could have a cultural reference and context that could be lost but ChatGPT was on point. I think if you are experimenting with ChatGPT, I would suggest that you explore anything that comes to your mind. It is amazing what you can learn from it, what you can walk away from. Now, on another note, I wanted to ask ChatGPT what it had on me. And so I did, I asked ChatGPT, "Who is Zohra Mutabanna?" There was no response. Well, I asked have you heard of Inside Tech Comm. ChatGPT confidently responded:"Yes, I am familiar with the show Inside Tech Comm. It is a web series produced by Adobe Tech Comm". I was not happy with that response. obviously because I own the show. So I tried to be more precise. "Have you heard of the show Inside Tech Comm with Zohra Mutabanna?" ChadGPT responded: "I apologize for the confusion earlier as of my knowledge cutoff date of September 2021. I did not have any information about a person named Zohra Mutabanna hosting or appearing on the show Inside Tech Comm. However, it is possible that new episodes or seasons have been released since then, that I'm not aware of. Inside Tech Comm is produced by Adobe Tech Comm and features various experts in the field of technical communication. If Zohra Mutabanna has appeared on the show, she may be a technical communication expert or practitioner who has shared her insights on the series". Now, thanks to ChatGPT, it is confirmed I have a long ways to go before I'm well known. But what what did I walk away from this response from ChatGPT? It had no information about Inside Tech Comm, but it laid claims to the fact that it's produced by Adobe Tech Comm. That's not based on facts. So you have to be aware of the fact that chat GPD may give you a response to a prompt, that is not accurate. The onus is on us to make sure that we verify and we back it up with with known sources, and not just believe what ChatGPT spits out, right? Also, despite the fact that my show has been around since 2021, its data set that it is reading from seems to be limited. We cannot trust the data it provides. In a recent CNN article that I was reading, titled AI and the future of work, the article calls out that the onus will have to be on us users to do everything from double checking the accuracy, to changing the language, to reflecting the tone we want. It will also be important to get buy-in and support across workplaces for the tools to take off. Basically, once AI takes off an AI is now being included and leveraged in Microsoft and many other technologies that we use at work. So it is very, very important to keep these points in mind and remember that the onus is on us. Now, what are some pitfalls to consider as AI permeates every facet of our life. These are these are just some of the pitfalls that came to my mind. And these are just my musings and that I'm sharing with you. What happens if there isn't a generative AI outage? Right? What happens if Chad CPT is down? It has been down. But what happens as we start using this technology on a day to day basis, and we start depending on it more and more. I don't have an answer to it. But these are just some questions that I'm posing to you that we need to think about. What are some unknowns as we use this technology at scale? We don't know what those unknown are today. What could this technology balloon to and what would happen to our jobs. I mean, that's a very real threat to our jobs. But we we what what can we do today to to evolve with it. But as we evolve with it, we want to make sure that we become aware of the unknowns as they read them selves to us, and prepare as much as we can, knowing what we know right now, as we use this technology at scale to be to be mindful, I think that that is the that's the best thing that we can do at this point, right? And to prepare ourselves and to and to make sure that we are using it, not running away from it. Another thing that I've been reading about is as we use artificial intelligence to assist us, it'll help us to do more high level thinking sort of work. But what happens? Will we lose any cognitive skills? As we give up on the grunt work, something to think about? As we start depending on this technology, as we use this technology to brainstorm, are we alienating ourselves from each other? Will there be a rise in mental health challenges as we isolate ourselves from human interaction? And then a very, very valid threat is data privacy. In an article published by Engadget, there was a bug in ChatGPT that happened in March. where ChatGPT accidentally revealed user chat histories. Now, that is not a good thing, right? So we definitely want to think about what happens as we start using this technology. Now, about a year ago, I was interviewed by Technical Writer HQ to share my thoughts on will AI replace technical writers. I will share the link in my show notes if you're interested in listening to what I shared at the time. But one of the things that I had suggested was how we need to stay ahead and stay abreast. Little did I know that ChatGPT was in the works. Now that the technology is here, I would like to suggest let's partner up with it. I'm reading some extreme opinions on LinkedIn, about how people are averse to this technology and we'll never use it. In my humble opinion, this will only work against us. Let us embrace it. And let us empower ourselves. On a final note, I tried to plug in as much as I can to hear more about ChatGPT. To learn more about it. There was one such webinar just about about two weeks ago that was hosted by the North Texas Lone Star chapter of STC, presented by Ryan Boettger. He calls himself as a techno optimist. And this gentleman believes that prompt engineering will be an emerging field in the near future. So guess what, I'm going to be focusing on prompt engineering. I can see its potential right. He also talks about taking on a mindset of artificial intelligence as an augmentative and assistive technology. Great advice. I will sign off on that note and I hope you will think about how you can leverage ChatGPT and any other AI technology out there for your benefit and work with it. Good luck and catch you on another episode.