ChatGPT and We - podcast episode cover

ChatGPT and We

Oct 24, 202322 minSeason 2Ep. 81
--:--
--:--
Listen in podcast apps:

Episode description

Rich and Paul struggle with video technology as they seek to jam the podcast into YouTube, but limp across the line with a whole podcast. Briefly digressing and explaining themselves, they proceed to discuss AI, namely sexy robots (they're against), using a complicated metaphor involving WordPerfect 5.1—definitely something Gen Z will connect with—and what's good about ChatGPT. Rich closes it by predicting an almost inevitable nightmare scenario where giant AI platform companies will use their advantages to destroy society without even trying. Listen to find out how! Good times! Brought to you by Aboard.

Transcript

is Rich Ziyadi. And I'm Paul Ford. And you are watching the Ziyadi and oh, damn it, damn it, damn it, damn it. It's the aboard podcast. So Rich, what's your job? I'm the co-founder and CEO of a board. You report to me, Paul. It's just fantastic. I am the president and co-founder of a board. And we've been working together for 10 years. We've started businesses of our own multiple businesses now. And we also have helped a lot of other people start through businesses.

So in case you're coming to this the first time, Rich and I had a podcast after we sold our last business, I called the Ziyadi and Ford podcast. It's a good looking photo there. Yeah, we got, for those of you who are listening, we got a little YouTube action going out. We're talking, we're looking at the camera. Okay, so let's clarify. First of all, we're not going to just shell a product every podcast. We're going to still talk about it. Every other podcast. Tech, life, business.

We just had this product to sell things. We were doing our podcast and we were, oh, and it has a sponsor in the sponsors of our company. And it got a little ridiculous. So we flipped that. We're now going to talk about technology and culture and the context of our tool that is designed to help people keep track of technology and culture and links and do work. And we're going to show it to that side. Anyway, anyway. I think the other thing worth sharing is that we are now also on YouTube.

Yes. So you can see pretty pictures and our faces along with our voices. If you want to keep listening, just audio, totally fine. Might even recommend it. Yeah, might be better. So we are going to look into the future with this podcast, Paul. Great. What are we going to talk about, Richard? We're going to talk about Word Perfect. Oh, that was a good word processor. You've got one minute. What is Word Perfect? OK, one minute. If I talk really fast, that's about 600 words.

So or two pages of use, Word Perfect. Use the Word Counting tool. Let me take a step back. Word Perfect was a boring looking, but very powerful, early word processor that let you make documents with your very, very goofy, dust machine. Like it had as much power as 1,000 of your watch. But boy, could it make a pretty page? It asked you had to learn. It had a lot of complicated commands. All that for, baby.

All that for, I mean, and so you needed to kind of gain some expertise to get the value out of a word processor like Word Perth. Sure. Picking up and playing, phones have made it so easy to do anything anymore. Word Perfect required some experts. Nothing's going to convey to people that were cool hip startup founders more than us talking about Word Perfect 5.1 for DOS.

There was a day where you had an overlay on your keyboard that had all the like different key sequences and key combinations to do anything in Word Perfect. And look, that's what takes us so long to get dates in college. I don't know. And we keep going. Yeah. So, then things got a little bit better. Yeah, Windows shows up and boy, is it pretty? And you know the date? It's not actually, it actually looks ridiculous.

It looks ridiculous, but there are little pictures of things that represent stuff. And there's a mouse, a mouse is this effectively a new interface paradigm where you didn't have to memorize key commands, you could point a little arrow at a thing and click on. Can I talk for two seconds about what's really happening here?

Yeah, absolutely. So, Word Perfect shows up and it's kind of like, yes, it shows you the character's on screen and there's kind of a loop running and it's waiting for you to type things in. But it's still very command driven. Like, you're going to give it a command and then it will do something to the document.

Windows is this like two-dimensional, brightly colored space that you navigate with a pointer and you sort of interact with these objects and every character on the screen is its own kind of object, and I'm different fonts and all this kind of stuff. So, it was a, it's not like this is like a little change. This is a vast change. Well, I think to me, it's like, it's going from computers being accessible to a million people, to a hundred million people.

Like, you could, at the child could pick up a mouse and move it around. It immediately gets what's going on. That's totally real. Right? And so, you see over time this evolution of how people interact with machines to get things done or to play. And now we are in the era of AI. Paul, do you know what I love? What do you love? I love weird, creepy robot like human or human like robot people holding their other robots. Fancy gin or golf. That probably exists. Excuse me.

You could use AI, we're going to talk about AI, to generate just about any kind of robot you like. Oh, that's pretty cool. So you know, when you look at, I do like these marketing images where people go out and they always have to make the robots sexy. Like it can never just be like a nice robot that's like, you know, it's creepy. It's pretty creepy. It's sexy robots, big packs. Yeah. And they're always really thoughtful.

Yeah. Yeah. The history of tech is the history of just amazing, amazing innovation after amazing innovation. And images of extremely horny robots. And that's who? Yeah. The last few years, he's had some false starts here. I'm going to throw out some words and I want you to react to them. I'm excited to react. Crypto.

Yeah, crypto is about the idea that when people go and do things like look at pictures or buy shoes, what they want is a fully fledged decentralized marketplace between them and the object of their desire. And that is the creepiest definition of crypto. It didn't take over the world. Banks are not under threat. You select to buy your house the old fashioned way. I mean, the three people are listening to this kind of thing. Well, Bitcoin's up today. Good for you, Chum. But yes, no, that's real.

It's going to be around forever like COVID and herpes. It didn't turn the world upside down. The metaverse. This was crypto, I think actually genuinely was innovative and weird, even if it didn't kind of make cultural sense in the way that people expected it would. The metaverse is just like, we got 3D and we don't not have enough stuff on our roadmap. We can't actually breed people fast enough to get them into the social networks at the rate that we need to justify our valuations.

Can you try it if you've gotten into the metaverse? I've done all this stuff, man. Well, I don't have the goggles. I've looked at all of it. I used to, I used second life a couple times back then. Is it fair to say that the hype and just all the frenzy has kind of really died down with it? I mean, the metaverse is if you're talking about quality duty or it's doing great. You know, Hilo used to do great.

But people don't want to do this idea that you're going to be able to create an abstract space without any kind of motivation aside from your own. This is a struggle. You don't want to buy the digital scarf at the digital mall. Here's what happens, dude. Yeah, you know, you know what they call that? Remember when they were always like, you'll be able to buy the sweater on the street like a watch TV? I think they actually call it like Rachel Sweater from Friends. That's how old the idea is.

It's like from the 90s. Interesting. Yeah. So that didn't take no. They'd be bundled together and said, well, crypto and metaverse is all the same thing and that's Web 3. Let me tell you what happened there. Look, I'll take it out of step. You and I, I'm the president of the year of the CEO. You know the number one problem we face as a company? What's that? I mean, lack of revenue because we're a startup. But like, it's that we continue to put our own needs before the needs of the user.

We do this all the time. We talk about like, we need more this. We need more that as opposed to, we need to deliver value to people so that they remember and connect to us. Yes. Okay. That is what the metaverse was. The metaverse was giant tech platforms assuming that their needs were more important than the needs of everybody else. All right. The people of Stats here. I do not live up to the promise. I've seen things by the way that announced they lurch forward.

It kind of is putt seen, clumsy and it doesn't take. And then it took like 10 years later. So who knows? I know. I hope those chief metaverse officers who were all named about three years ago are doing great. I bet those LinkedIn. I bet there's a strike through. Now we are. With the guy, there's chat, GPT. We've got mid journey, essentially airbrushing the side of my van every day. Every day. Put those dragons in there. Another false start. Another false promise? No, not at all.

Culturally it's gone banana cakes. That's things tend to do these days. So like, you know, dudes with beards saying that we are going to have, you know, artificial general intelligence and we're all going to be enslaved by the robots and they're kind of into it. And any of other people saying that like, this is the absolute worst thing that could ever happen to culture on top of all the other worst things that have happened to culture.

So like, we're, take the whole conversation of it at a side like, it's really interesting technology. I'm going to not be the person to advocate for it. Bring up, actually bring up our product. We keep track of the links for this thing and a board for the, there it is, sexy robot. So rich. Yeah. So you're impressed. I am impressed. I think that this is, there, I don't want to be the advocate here because actually someone did a much better job than I did.

There's an open source and just kind of a general good actor in the world of technology of prison name Simon Willison. There were that Simon. Their website is SimonWillison.net and they gave a talk recently, he gave a talk recently, making large language models work for you. And it's online. Yeah, it was at the word camp at the big WordPress conference. Okay. It's online. And the primer. It's been an hour long. It's excellent.

And Simon is someone who just goes in deep on things and he's been really obsessed with this space because he describes it and I think very well is almost like an alien technology. This thing has landed in our world and it doesn't work like any other kind of more algorithmically driven tools that we've used before. And the behavior is different. It's different kind of technology. It's sort of like a wild drunken intern.

It will just give you something no matter what you ask, but very often when you ask for something, it also behaves like someone's typing back to you with some far off list. Yeah, it's just a very strange new way of structuring data and embeddings and vectors. And certainly in all these other words that sort of get thrown out.

So Simon has gone in deep and explained sort of the utility of it, how to make use of it, how to explore it, and sort of also a little bit, not the ethics of like AI in general, but the ethics of how to use it. Like if you're going to create an image with AI, share your prompt and show, show, you know, what is a way to be an ethical consumer and who applies this and uses it, you know, a lot of it comes down to citation, make sure you explain to people where the AI is involved or where it isn't.

All right, I want to go slightly abstract here, but I want to make a point. I love a good abstraction. I know, I look at computers as almost like this relationship between humans and machines. I mean, I don't, wow, that's extremely, wow, did you come up with that? That's amazing. Let's keep going. We talked about one perfect before. Yeah, we did. In the early days, computers were really impressive, right? But it required work. They've seen the impressive at the time.

Of course, now they look like ridiculous jokes. Also, it was on their terms. We have to learn the tools. Yes. You have to educate yourself. Otherwise, we're going to get the value. But this thing could add up a bunch of numbers and then it could go, and man, that seemed amazing. Then the mouse pointer comes out and then the windows come out and what ends up happening is that a lot less is asked of us. Yes. You don't have to be an expert anymore. Yes, you don't.

In fact, you can shoot a movie on your phone. It's similar. I would analogize it like you go, you used to have to follow a recipe and now you can go to the supermarket and get a link. To also write to the microwave. Get a lean cuisine. Yeah, exactly. And that's amazing. That's what progress is when you reduce the barrier to entry, more people can use the thing. Yes. More people can enjoy it. You have to be careful with kids, computer phones and are a little scary.

But yes, a kid can pick up a phone. You watch a three-year-old on an iPad. It is something to see their hypnotized pretty much. And we have unlimited access to most of the world's information wherever we go, which is long been a dream of humankind. Here we are, right? And so we are in a place where computers have outpaced what's asked of us. They know that sounds weird. No, that's right. They can lean in further than we can. They can. And then in the 2000s, we could start to talk to them.

Yeah. So it's like, hey Siri, when was Napoleon born? And it'll tell me. It'll tell me it's a different one. No, it'll tell you that Napoleon is a pastry. But yes, I got it. It messes up sometimes, but we're very forgiving as a people, right? Oh, yeah. Now we're at this fascinating point where the AI wave has kind of sprinted way ahead of us as humans. Look, here's the thing. Technological innovation evolves more quickly than humans do. Well, no, actually. I can say that with real accuracy.

We sure have. Those humans don't evolve over, they evolve over millions of years. A lot of them have to die. So the, no. But let's not, I wouldn't put it that way. I would say that humans, anthropomorphize and find new things very sort of mystical when they're actually practical. So new technologies show up and people tend to, and there's a famous Arthur C. Clark quote, which is any sufficiently advanced technology is indistinguishable from magic, right?

I would actually say that we could amend that for 2023. Any new technology is essentially indistinguishable from people. We tend to, we project and project and project. And I do think over the next few. Well, I mean, look, the products are getting names. This is serious. This is actually, what's the guy's name and Microsoft? Cortana. Clyde? Clyde will make you a sandwich. Clyde's like, you want some so long. You want some so long. The only American and Mayo. White bread.

Simon makes a really good point, which is like one of the kind of ethical failures is not, there's all kinds of issues, right? But constantly anthropomorphizing is really bad for us as consumers, because the robot keeps telling you it's your friend. Yeah, hold on. Give me a second and I'll get you the answer. Hi, it's me. Sky. Remember when they tried to date as a Butler? Yeah, jeeps and stuff. Or you know, that's the thing, right? Or a clippy.

And they keep making it into a person as opposed to like a set of data things that are interacting. So we learn less and less about what's going on under the hood and fantasize more and more that it's like our thing, our person. I want to glimpse into the future though. I would argue that today it's actually quite archaic and the AI brain is quite slow. It can only be one thing at a time. It's still batch mood.

Like it's still I put the punch cards in and then it comes back when after processing. Imagine a world poll. You bought a bicycle. Okay. You call Chachi PT, Sky. Hi Sky. Hey Sky, just bought a bike. Good for you, Paul. Thank you. Thanks, guys. Because your lower body is a disaster. I'd like to see your legs. I'll talk to you later. It's like two fish bellies. Stay with me, Sky. Okay. It's in pieces. I need your help putting it together.

Wow, that's great because I know exactly how to put a bicycle together. She tells you to flip your camera. Lay out all your parts, Paul. The way you said that actually makes it sound weird. But, okay. Bicycle. That actually sounds way more realistic as to how people want to use this thing. Okay. So, she's just going to say, hey, that blue screw. It's very awful. That blue screw, you know, where's your screwdriver? Where is your, okay.

Yeah. Yeah. Paul, you don't have all the tools to put this together. Let's go to the hardware store. I'll pull you back. You go to the hardware store now. You call her back. I got this wrong. She's going to say, I'm going to order you a really amazing screwdriver. Paul, Amazon's in the mix now. And then the drone is going to drop it on your head. Yeah, I got this buzzing in.

What I'm trying to get out here is that the innovation ahead now is the interface for humans to talk to these machines and the machines being able to process. No, I mean. If you talk about a video call coming out in the AI engine. Oh, good, dude. That's a lot of data. You know, it's true. Okay, so what's going on is like, it's good. It takes the static image. It breaks it apart and it figures out what's going on inside. You can't pan the camera around. Right? That's huge. That's huge.

We're having a conversation right now, but you know, like that it can't actually have a conversation still kind of batch mode, right? In a way, I guess what I'm trying to say is you think, oh my God, the machines are coming for us. But my God, the human brain and its ability to work in real time is something else. For most of the people I've got, you know, it's not all. You know what I can't stop thinking about that?

It's going to, like, Ikea instructions, like, that those terrible manuals that are just like no language. Yeah. I think I'm imagining that perjury of this world and it's just like, hmm, yeah. Let me close it with a question. It should, no, she's just, Skye is going to have you, you're going to try to put together an expedient, hmm, hmm, hmm, yeah. It's, you know, now, now, now a lot of disapproval. And there's a little froggy people and it's like, you need a friend to help you lift it up.

Is Skye going to call you for data? Hey, Paul, I saw you wrote an article about Napoleon on Medium. I have a question for you. So Skye, this isn't the best time right now. I, this is terrifying. What? You only want. Want, want. You need knowledge. Yes, she can go get it off the internet. No, because she knows you're an expert on Napoleon. Skye, why can't she call you? Because she's going to call you because they can either just put your phone up to your neighbors window for a minute.

You know, somebody's really curious. I've got it. There's some data. Yeah, I don't, I, like, that, yeah, you don't want to be a conduit for knowledge. It's not to say humans are very dangerous. Like, we have a lot of systems like universities and, and journalism, like, I'm talking about like meta systems that we can, that, they've trust built into. Yes, where there are ethical guidelines. So the ethical guideline inherent in something like Google, right?

Is it like, if it's published publicly on the web, Google will spider it. Yeah. And the law reflects that. If you have this ad hoc data collection platform that pretends to be human, yeah, that is a nightmare for our society. It can be a problem. And you got there. You got to the actual, like, this is, I don't think the fear, real fear. I don't think the robots are going to get super smart and suddenly take the whole of their missiles out of it. Shoot missiles.

I do think the robots are going to call us and say, did you know that Donald Trump wants to give you $25, right? Yeah. Hey, how's your son? Yeah. Yeah. And you'll be like, what are you talking about? And we saw glimpses of this with how smart the, you know, the Facebook ads were doing elections and how they were manipulated. This is a lot of power. Many, many people are deeply vulnerable to this manipulation.

Yes. And we do not have an ethical system in place in our society where everyone becomes an information gatherer. That's right. So that's right. Good. Good. Great discussion, Paul. I want to talk about a board. That's right. So earlier, we were talking about that. We use it to track the links for the podcast. That's one of the tools to rich and I are able to collaborate. We make a board and it turns the links into pretty cards. And we can kind of get our podcast together that way.

A board is a general purpose, data tool for all kinds of things. You can use it to go find an apartment with your spouse or you can use it to save the TV shows you want to watch. It's our attempt to make a data and software unbelievably simple for everybody. Everybody, everybody, everybody. Yeah. So we'd love for you to try it out. Let us know what you think. Do it all board.com. Click the join button. It's free. What else do we need to know? Why would pass? We just want you to like us.

Oh God, just like us. Follows up on YouTube, five stars somewhere else. Five bars? I don't know. No, seriously, I've turned my deaf for a new approval. It's a whole career. So all I need is for you to click that little bell. Please, please, please God. I think this one pretty well, Paul. We've recorded this particular podcast seven times. Yeah, let's see if it actually recorded. Yes. I have to glue together all the various pieces. But let's see how it fits together. Have a lovely week. Love you.

This transcript was generated by Metacast using AI and may contain inaccuracies. Learn more about transcripts.