ChatGPT, Lensa and DALL-E are giving more people without computing skills the chance to interact with artificial intelligence. These AI programs that can write detailed responses to questions or create stunning images based on prompts are all over social media.
WSJ personal tech reporter Ann-Marie Alcántara joins host Zoe Thomas to discuss how the programs work and the concerns that have been raised about their potential misuse.
Photo: NurPhoto / Contributor
#ChatGPT #AI #WSJ
Foreign [Music] Briefing for Wednesday December 7th I’m Zoe Thomas for The Wall Street Journal Today we are talking about chat GPT a State-of-the-art conversational AI model Developed by openai chat GPT is capable Of engaging in human-like conversations Providing answers to questions and even Generating original content That little introduction was written by Chat GPT itself after just a short Prompt from Maine the artificial Intelligence chat bot is suddenly Everywhere reaching a million users on Monday just a week after it was released According to its maker open AI Openai also created Dolly an AI platform That creates images including some Award-winning images based on users Prompts and those aren’t the only AI Programs becoming internet Sensations But the popularity and sophistication of These programs also raises questions About how they could be used or misused Joining us to dive into that is our Personal Tech reporter and Maria Alcantara hi Anne-Marie thanks for Joining us Hey Zoe thanks as always for Having me so let’s start talking about Chat GPT can you just describe what it Is for any listeners you haven’t come Across it yet so first of all chat gbt Is a tongue twister as you can tell very
Hard to say uh really fast but most Importantly it’s a free AI chat bot Really you can ask it anything and Everything you can ask it you know basic Questions that you would Google too even Asking it to write a love song for you And it’ll spit out an answer and you’ll Get a different response every time you Ask something Okay so for anyone who hasn’t tried it Out yet can you tell us what you know Some prompts people are putting in and The kind of answers that chat GPT is Generating Yes so you can ask it really any type of Question you want as for example we’ve Asked it if unicorns exist and it said You know a very nice little answer that Unicorns are a fictional character are Most likely don’t exist others have Asked it you know really deep questions Such as you know when will I die and It’ll tell you it can’t answer for you Because it just doesn’t make those types Of predictions and other questions that People have asked also include very Existential ones such as you know who am I and the chatbot will respond and tell You it can’t answer that for you and Like only you can Define yourself and Who you are so it can be deep it can be Funny it can be serious it can really be Anything you want it to be So those are some like really
Existential questions what about some of The more creative prompts that people Are putting in yeah so the other day I Asked the chatbot to write me an essay About quantum physics and Pokemon two Very unrelated topics and the chatbot Spit out a great essay about what Quantum physics is what Pokemon is and How they actually relate to each other Even though one is rooted in reality and The other one’s you know a fictional Character and so people have been asking It questions like that people have tried To get it to solve homework questions Engineers who work in computer science Have also tried to you know figure out Problems and solutions that they’ve been Having in their work day and finding Responses to that as well How is it coming up with these answers So chat GPT was trained on massive Amounts of data that the company has Gathered from the internet and other Sources through 2021 and as you can Imagine that’s just it’s infinite you Know what doesn’t the internet have out There Wikipedia alone has so much Information right so it’s kind of Trained itself on that information and Pulling from that to answer your Questions are there limitations to these Answers there are some limitations like I described sometimes it won’t tell you A response because it can’t or you know
It’s too essential some of the other Limitations people have come across is That the chat bot will tell you Information that isn’t correct and you Know maybe you’re a scholar and you know That for a fact let’s talk a little more Broadly about AI software because chat GPT isn’t the only one that’s taking off What else is out there and what is Behind all of its popularity Some other interesting AI tools that Have popped up include lenza which is This app that is at the top of the App Store charts right now and it sort of Takes photos that you feed it and makes A beautiful artistic photo some of it’s Like really space like some of it feels Like it’s out of Middle Earth another One that went popular a few months ago Was Dolly 2 which is a similar function To lensa where it takes things you’ve Written down like podcaster in space Recording a session and it’ll turn that Into an image and really what’s driving All of this is that these AI tools are Now being Given to the public given to the Everyday person and you don’t need to Have a degree in advanced you know Artificial intelligence or computer Science to know how to use them you can Just do exactly what we’re describing Put in whatever prompt you want and see What the results are
I mean if everybody has access to these It does raise a lot of questions about How it could be used I mean difficulties Say for for teachers if students use These to write essays or you know even For people making a living for artists Or you know journalists trying to write Stories Those concerns have definitely been Raised with all of these types of Products and many AI experts that I’ve Spoken to have said that we should view Some of these tools not as a goodbye to Those roles and professions but more so As an enabler of enhancing your work so If you have writer’s block You could type in a prompt and see what It tells you and then maybe that’ll you Know get the juices flowing in your own Brain naturally if you’re a student Maybe it’ll help you understand the Question better and come to your own Conclusions and answers of course people Will probably use it in ways they Shouldn’t but we’ll see how what you Know guard rails end up being put in Place to make sure that these things Aren’t abused in the ways that they Definitely could be All right that’s our personal Tech Reporter Anne Muriel Cantera thanks so Much for joining us thanks and that’s it For today’s Tech news briefing for more Tech stories check out our website
Wsj.com I’m Zoe Thomas for The Wall Street Journal Thanks for listening [Music] [Music]