Chat GPT, an Artificial Intelligence chat bot, launched in November 2022, and now, it’s everywhere. Each day I read something about Chat GPT in my news feed or hear hosts discussing it on podcasts. Tech enthusiasts revel in its possibilities. Teachers fear it will promote cheating. Creatives worry it will steal their jobs. But what is it?
What is generative AI?
Artificial intelligence has been around for a while, and most of us use it every day. Do you talk to Siri or customer service robots? Then you use AI. In simple terms, traditional AI analyzes large amounts of data to find patterns and make predictions. The new trend in AI is generative AI, which is what ChatGPT is. Generative AI collects tons of data inputs, such as words, descriptions, images, or computer code, and creates something “novel” in response to a prompt. For example, one could ask a generative AI model to create a painting of a dog in the style of Van Gogh, and it would create something that looks like a Van Gogh painting. But isn’t.
What is ChatGPT?
ChatGPT is a language processing model that provides “human like” text responses to prompts. For example, you can ask it to write a poem about cats or an essay about World War II. Chat GPT is run by Open AI, a research lab backed by Microsoft. ChatGPT is a form of generative AI.
Generative AI models have been around for a while, but lately they are getting more attention because Chat GPT is free and open to the public. All you need to use it is an account with Open AI and ChatGPT.
What are the benefits of ChatGPT?
Open AI says ChatGPT has huge potential for more efficient work. It could help content creators by providing tips on marketing strategies and copy for websites, blogs, and social media posts. ChatGPT can take over tedious tasks and open up more time for employees to work on higher level thinking, like research and strategy.
Some argue ChatGPT can do some basic tasks better than humans can, because it sees patterns humans do not. For example, ChatGPT might do a better job of writing copy for a website to maximize search engine optimization. It can also, ironically, do a better job of personalizing text.
Another benefit: improved internet searches. Some believe a model like ChatGPT will eventually replace search engines like Google. The generative AI can give users actual information, not just links to websites.
And on the coding front, ChatGPT can make basic coding faster. In fact, Microsoft is putting a lot of money into developing generative AI because it wants to use AI tools as “pair programmers” to improve coding efficiency. AI can often debug code faster than humans can. I bet my son, who recently spent six hours debugging a program at work, would pay attention to that.
Limitations of ChatGPT and other AI models
ChatGPT has its issues. To work well, generative AI models require millions of data points and significant computing power. They also need human input at the beginning and end of each process. For example, a person has to give ChatGPT a prompt so it can generate a response. Then, the person has to evaluate the response, and if necessary, provide new prompts to refine the final output. Because generative AI depends on data to produce things, it has inherent bias. It learns from the information you give it. Even if it’s bad information.
Open AI admits ChatGPT is not good at math calculations. For example, if you tell ChatGPT that 10 + 9 = 20, it will carry that incorrect information forward during your chat with it.
Also, ChatGPT is trained on data up to 2021, so you cannot use it to generate up to date text about breaking news or current events. Also, since it is easy to manipulate, it may produce offensive, biased responses and misinformation.
What to watch for with ChatGPT
Like all technology, ChatGPT offers benefits and drawbacks. As a creative writer, I join the ranks of artists, journalists, and creatives who worry AI generated content may replace the work I do. However, ChatGPT is good at filler language and basic document generation, but it can’t write a novel. I’ve read some songs and poems written by ChatGPT. They are not good. So hopefully AI won’t take over the author world.
Educators worry ChatGPT will make it easier to cheat. Students can ask GPT to write an essay on The Great Gatsby, for example. Or to write a paragraph explaining the Krebs Cycle for biology class. Universities are scrambling to respond to the implications of ChatGPT. Some professors are contemplating oral exams and in class writing assignments, with, dare I say it, pen and paper! (This old Word Nerd thinks that’s a good idea with or without ChatGPT.)
One of my biggest concerns is regulation of AI. Or the lack thereof. Currently, artists are suing Microsoft and other tech companies over AI models. The artists say AI models scrape images and text from the internet to create items that are basically collages of original work. The AI models do not ask for permission to use the copyrighted material, nor is AI currently required to follow any licensing guidelines. The artists say that’s intellectual property theft, and I agree.
AI can blur the lines between what is human made and what is computer made. We already have trouble figuring out if what we read on the internet is true. Generative AI has just complicated that problem.
And what about accountability and responsibility? If a generative AI model produces something offensive, inaccurate, or slanderous, and it goes viral, who is to blame? The person who prompted AI to create the offensive material, or the company that runs the AI model?
We do not have legal and ethical standards established to govern AI and its use, and we need them. Quickly. You may be thrilled with the possibility of AI. It may scare you. But we all should be paying attention. Generative AI is here, and we need to make sure we use it wisely.
What are your thoughts on ChatGPT? How are you staying informed?
Thanks for getting nerdy with me!
Julia,
As usual, I am delighted to read this and learn from a trusted source what this new technology is about.
Thanks,
Jane
Thanks Jane!!
Technology is moving much faster than we can make laws to regulate its use. However, the past has shown us that, after an initial wobbling, we learn how to assimilate the new technology in a way that is just and helpful, or at least as just and helpful as we humans can be.
Every advance is a two-edged sword. I recall that several years ago, I read an article in a Life magazine from 1961. The writer was a futurist, someone who tries to figure out what will come to fruition in the next generation or two. I do not know if such futurists are so popular now, but they were back then. Most of what I read of their predictions never came to pass or only in a very vague way – flying cars are still not ubiquitous! But this particular futurist nailed it on what is likely the most influential technological advance of our time, maybe of the last two millennia – the internet. He said that in the not-too-distant future, nearly everyone would have a computer in their home and there would be a system in place that would enable them to all be networked which would result in an explosion of information as it would be difficult to regulate or censor what was transmitted over that network.
He went on to say that this would also result in an explosion of pornography. When you consider that, in 1961, the only widely available pornography was Playboy magazine, and at that time it was relatively tame compared to what would be published 15 or so years later, it was truly prescient for him to foresee this.
Of course, he was right – nearly everyone has some sort of computer in their house, and nearly everyone is connected to the internet – and this has resulted in an explosion of pornography. This is certainly a downside to the internet, but I do not think that it outweighs the benefit we gain from the internet. After all, were it not for the internet, I would likely not know that you exist, much less would I be able to benefit from what you write.
The music and movie industries were greatly affected by the advent of the internet as it became easy for people to “pirate” recordings. It did hurt these industries for a few years, but they learned to follow the proverb, “If you can’t beat ’em, join ’em,” and began making their content available through streaming services and are now making much more money than they ever had made before.
In truth, I do not know when computer programming passes from mere programming to AI. As near as I can tell, it is all just varying levels of sophistication. And there will be some casualties along the way, but we will learn to adapt (it’s what we humans are good at) and will learn to make good and just use of it.
But of course, by the time we do, there will be even more sophisticated technology which we will need to learn to adapt to. There will always be a new steam drill, and there will always be John Henry’s who will vainly try to defeat them and “die” in the process. But the result of that will be an improvement of life for many more as humans are relieved a more and more of the “grunt” work and left freer to do that which is distinctly human.
At least I hope so! 🙂
I like your optimism! You give great examples, and I hope your futurist prediction comes true!
I just heard on a podcast that ChatGPT has acquired over 30 million users in just the two months since it has been released. In comparison, it took Instagram a year to get 10 million users when it first launched. So we are dealing with a difference in scale. Technology and innovation are moving faster than they did 50 years ago, and their speed will continue to increase. That makes it hard to see and prevent potential problems. We need to be mindful of how we use new tech. Not fearful, but mindful.
There are only 8 billion people on the world. Maybe the statistic means it has been used that many times.
I like your “not fearful but mindful.” That is a good attitude always to have.
Whoops! My typo. I meant million and changed the comment. Thank you for correcting me. 🙂
Well, I had to try it. I asked it to write the preamble to our constitution in iambic pentameter. I got this:
“In Union’s cause, we People of these Lands,
Do ordain this Constitution with our hands.
To form a state of Justice, Peace, and might,
Where Domestic Tranquility shall shine so bright.
For common Defense, we pledge our lives and all,
And for the general Welfare, we do call.
That Liberty’s blessings, for all time, shall stay,
For us, our Posterity, and all Mankind to weigh.”
However, this still leaves open the question as to whether it is our founding fathers or Shakespeare who have now been rendered irrelevant! LOL!!
🙂 Thank you for providing this sample.