Mary Daphne: Hello, advanced English learners. Welcome back to another conversation. I'm joined of course, by the one and only Greg. Hello. Hello. I'm excited for this conversation. What about you, Greg? Very excited. All right, so without further ado, let's get into the topic. You can explain it, I think.
Greg: Sure. Today, yeah. We are talking about a type of AI that's being developed, and what's really cool about this AI is it's something that we can use in our everyday lives. So it's not, some in distant future only super powerful companies can use.
Everyone can use. And the technology currently is called G P T three. Yes. Which I think we've mentioned in previous podcasts. And it's offered by a company called Open AI. So that's what we're going to be discussing.
Mary Daphne: Amazing. Let's get right to it then.
All right. G P T three.
Greg: Yes. And don't ask me what it stands for. I should have looked it up before we chatted. It is a product offered by Open AI. And open AI is this consortium of AI researchers funded by, some hot shots in , Silicon Valley.
Mary Daphne: That's a good way of describing them.
Greg: Yeah. And the idea here is that they are supposed to be responsibly developing artificial intelligence technology for the benefit of everyone. And that's why it's called "open." Now in practice, for the most part, it's actually been very closed off.
Mary Daphne: Yeah. Not many people have heard about it.
Greg: They've licensed it to Microsoft. So I would consider it to be relatively captured as opposed to open. And by captured I mean it's controlled by some large centralized entities. Nonetheless, it has produced some really cool product. And to their credit, they have released some of them and aspects of some of them to the public for us to play around with.
Mary Daphne: So in a way to beta test maybe. We'll see how long this free trial runs. I'm really curious about, how accessible they make it at the get go.
Mary Daphne: But I think, we were having conversation earlier and I think you mentioned something like they can't, put a paywall up forever.
So let's talk a little bit about the applications for a technology like this, which I'm super excited about.
Greg: Totally. G P T three in general is a AI service that takes things you say and does things. In response.
Mary Daphne: Based on your input, it provides some output.
So we talked about this briefly when we brought up the idea of the Dall-E software that you plug in, certAIn parameters, say a dog eating pizza on the moon, and it will create different types of images to that nature.
Greg: Yeah. It'll literally show images of dogs eating pizza on the moon of all different types.
Some look pAInted, some look cartoonish, some look hyper real.
Mary Daphne: Yeah. Photorealistic meaning it looks like it could be a photograph, which is just so mind blowing.
Greg: Yeah. And so that's really cool. And for me, I was like, okay, that's the big development we're going to see in this year. Maybe it goes to video even cooler, but that's sort of it.
The most recent iteration has done, has provided something that to me feels like a true intelligence. The previous, what you were just discussing, this sort of text to images is really cool. That still feels like a machine.
Mary Daphne: Yeah.
Greg: What we now have is G P T Chat. And this chat function allows you to converse with the AI entity.
Mary Daphne: Oh my gosh, it's so crazy. There's so many movies like Space Odyssey is one of them. That was one of the pioneers, the trailblazers
Mary Daphne: In this space with Hal
Mary Daphne: who? I'm not going to give away spoilers.
Mary Daphne: But interesting. He has a huge, important role. Hal? Yes. He's AI.
Greg: He's like the brain of the spaceship.
Mary Daphne: And then more recently in a show we watched on hbo, raised by Wolves. , there's, in season two you see some type of AI that governs a, a community essentially. Totally.
Greg: And asks a lot of questions. Where does the line between human and AI, where is that drawn? It questions who should be holding the power.
And I think there's arguments for both.
Mary Daphne: whether AI should hold the power or humans.
Greg: Yes, exactly. But before we get to that, and I think that's a really good conversation. Yeah. , let's talk a little bit more about what this product does.
Mary Daphne: So essentially, I'm going to play around with it. I haven't yet. I'm excited to dive in. But I signed up and I'm going to do that. So we can probably have a, segue separate video to this afterwards. . But essentially there's a chat box like we know, and you plug in certain parameters, an input. So what could that be, Greg?
Greg: Anything. The input, you can talk totally conversationally so you don't have to speak in stilted language. You can speak as if you were talking to someone. So literally, this is what I could type begin, quote.
I would like to plan a party for my friends. It's a holiday party. I love decorations. But I have no idea what to do for food and drinks. Can you help me plan this party? And this thing will respond with an answer, a comprehensive answer for everything that you mentioned in that question.
Mary Daphne: Unreal.
It's just so cool. And I was even going to go out and say, why don't you theme it? Like you could have a holiday party that's space themed.
Greg: Totally .
Mary Daphne: Yeah. Let's just,
Greg: even better.
Mary Daphne: Why not?
Greg: The more specific your input is. The better result you're going to get from this machine. So that sample output could be like, here are three different possible party favors that you could offer.
Here are three different types of drinks you can make. The general theme, should probably be space suits. So have, make sure that people bring space suits. You can put, fluff on the ground moon dust.
Mary Daphne: Oh, so cool. Yeah.
Greg: All these things this thing will respond in this incredibly comprehensive way and it doesn't stop there.
So once it provides its response, it has a memory, right? So once you open up a conversation, it preserves the state, right? It's got a persistent state for the duration of the conversation. Which means you can ask follow up questions and you could say something like, okay, cool. I like the idea of space suits.
Tell me what's the best way to make a homemade space suit?
Mary Daphne: Ah, I love it. And they'll tell you aluminum foil and toilet paper rolls and who, I don't know
Mary Daphne: I'm making of this up, but whatever it is,
Greg: I think it would definitely choose those.
Mary Daphne: Yeah. Yeah.
Greg: I'm thinking like one of those glass pot tops for the cover to your
Mary Daphne: Oh, that's good.
Greg: Space suit. You know?.
Mary Daphne: And now I'm kind of intrigued. We have our holiday party coming up and
Greg: Yeah. Maybe we can put it to use,
Mary Daphne: our guests are going to be like, okay, Greg and MD have,
Mary Daphne: Gone off the deep end. But no, in all seriousness, there's so many cool applications for this, for both, like a personal type of thing, or more professional applications as well.
If you're, working on something where you have a term paper. Your professor asked for high quality sources and you only have one book, but you need to find so many more sources. Maybe that's an application for it and, AI could assist you in identifying those sources.
Greg: Yeah. You can almost think of it like a supercharged Google search. Because this is a Google search that interprets everything you're saying. So you could say find me five of the most cited papers, the, most citation papers on the state of AI, and it will produce a result of that, right?
Mary Daphne: Yeah.
Greg: If you search that on Google, you're going to get a bunch of AI papers, but not necessarily the ones that have the most citations. This program will produce the ones with the most citations, right?
Mary Daphne: Yeah. Yeah.
Greg: It's much more directed in what it does. It's like actually having a human that you're, interacting with, listening to what you're saying and going out and doing it as opposed to search feels one layer deep, maybe two layers deep,
Mary Daphne: and there's no real follow up.
Greg: Yeah. This feels five layers deep with all the follow up that you want.
Mary Daphne: It's incredible. And even if you don't have that specificity, you can still use this in something like, I'm thinking about redesigning my office, I'm wanting a warm, productive vibe. What are some choices? And then it might give you like five different examples.
Greg: Right and this is what's really cool. So let's use your example.
Mary Daphne: Yeah.
Greg: A warm, productive vibe, right? It produces five examples of how to achieve a warm, productive vibe. You can then take that description that it produced and feed it into something like stable diffusion, which is, an AI image generating machine.
So now stable diffusion takes that description and produces five actual renderings of this office space that you want designed. So with just two commands. Even one, if you just hook them up, you have gotten beautifully rendered, examples of how you can configure your office, which normally would take, three different people, three different experts.
A lot of money to produce. You produced and weeks, weeks of conversation
Mary Daphne: and a lot of back and forth,
Greg: Tons of back and forth
Mary Daphne: emails and calls and maybe in person visits and trying to find a designer. And designers are very expensive. So you know, you'll rack up a huge bill. Yeah. So
Greg: all that in three seconds.
Mary Daphne: It's amazing.
Greg: Think about what that does for productivity. What can we do with that kind of rapid prototyping, iteration and discovery, right? It's just mind boggling that we're entering a whole new paradigm. And this scares a lot of people, and we've talked about this before, because it feels like it's going to take jobs away
Mary Daphne: and make them redundant
Greg: and make them redundant but from my perspective, what it does is it just enhances people's ability to do their jobs.
Mary Daphne: A hundred percent. And for me too, I have that same perspective on it, Greg, because it frees up so much of the in between wasted moments and instead allows us to harness our creativity. And use AI enabled tech enabled, products to enhance our productivity and our experience.
Greg: Yeah. It's completely paradigm changing and it takes what would require an army, i e, a large group of people to accomplish and condenses it down into, just a few services. And so that one person can really do a huge amount of different tasks without much specialization without much capital.
That's so empowering for, I feel this viscerally because I don't have much capital, but I love starting businesses. I love building things. And to be able to do this at an even lower cost is so exciting for me.
Mary Daphne: Totally true. And I feel like the old paradigm was, oh, you need to hire a lot of people.
You need to do this, you need to get a team together. People got to be working for you and with you. Okay, maybe. Yeah. But. How cool would it be if AI could help with that as well?
Greg: Yeah, and then we can all do our own thing. Look, if you enjoy working for a larger group or with a larger group, that's great and go for it.
But a lot of us feel, I can certainly speak for myself in saying I don't really like working for a boss. I'd rather be my own boss, and managing a large team is stressful for me. So being able to leverage, technology like this to do a lot of the things that a team would normally. Is, I don't know for me it's very empowering.
Mary Daphne: Yeah, totally empowering, very liberating. And it gets you to think outside the box.
Mary Daphne: So really fascinating stuff. So I'm going to try it out. I encourage you to keep your mind open about this. I don't know, maybe you're already like super gung-ho and you're like, I want to try this.
I think this is super cool. What are some of the applications you see for your own personal use or your professional use, how do you see this kind of technology enhancing and empowering your life? Let us know in the comments and we always love hearing from you. Be sure to share the podcast. We are so grateful to all of you that have already shared it with people and keep doing that.
It's amazing. We appreciate all the reviews and the comments you write so keep doing that and don't forget to check out our blog as well. All right? So just go to advanced English dot co and we'll see you in the next one. Bye for now.