Buff Presidents, Google Will Crush ChatGPT? + AI Gordon Ramsay

AI4H Ep 12
===

Kevin: [00:00:00] can we use mid journey to make a 12 pack but of forehead muscles? Like you get so smart listening to this podcast that you get mental abs. 

Gavin: here we

Gavin: go. At 

Kevin: Episode 12, baby. Welcome 

Gavin: everybody to AI for Humans.

Gavin: Kevin, what do we have on today's 

Kevin: show? We have got incredible new imagery tools that will let you make video from me. Text what, what is this? 2023. Also, you can drag your photos and morph them into all sorts of weird digital nightmare fuel. We'll show you how. Google's deep mind claims that they're coming after Open AI and Chat G P T.

Kevin: They're going to crush the AI competition. We're gonna have more on that. And the Grimace Shake is not anything I had on my Bingo card for this year, but the Grimace, me either Shake has taken over TikTok and we have Gavin. A very [00:01:00] special guest. We've got a chef, Ooh, who do we have today? World famous presenter and an idiot sandwich enthusiast.

Kevin: Gordon Rams AI will be here to tell us all about the Krem shake. 

Gavin: \ Welcome to our show. AI for Humans is a exploration of ai. I would say it is a, it is a fascinating dive into the world of artificial intelligence and the tools and tips you need to survive.

Gavin: It is also eight grind core things that you can do to make money with Chachi pt, right, Kevin? 

Kevin: Oh, let me tell you why. Three AI experts absolutely hate us. They're nauseated by this podcast, one of my little treats. Every week, Gavin is watching you grab the machete and go whacking through the weeds to try to figure out how to succinctly describe whatever the hell this podcast is each time, because.

Kevin: I, I like, I see it. I see you recognize it. It's on the horizon. It's oh, where it's a broad discussion and celebration of ai. You usually [00:02:00] say that it's, what is it? AI for smart people who like dumb things. 

Gavin: Yes. That's, that's kind of the temporary slogan that we've come up 

Kevin: with our brand.

Kevin: We gotta codify this. Yeah. Because if I try, I wanna get it tattooed on my body, but I can't have it struck through every week and go, I will not have enough abdomen to have it completely written, written out English. Don't worry. In the future, 

Gavin: in the future, you can add abdomen that's coming. Sweet. The abdomen ad is coming.

Gavin: Yeah. It's a new AI drop that's coming next week. You've tried the 12 

Kevin: pack, but have you got the 24 pack?

Kevin: Add abs to your side. Add abs abs. Add abs to your back. Oh, now we gotta figure out abs on forehead, abs, abs on forehead. Can we use mid journey to make a 12 pack but of forehead muscles? Like you get so smart listening to this podcast that you get mental abs. 

Gavin: This is a great idea. Next week we will do mental, ab, forehead, abs.

Kevin: Have you flexed your brain today? AI for humans? Look, I got, I can [00:03:00] do 

Gavin: it myself. If you're on video, you can see my own mental abs. I've got them 

Kevin: already. Oh dude, he's on trend.

Kevin: He's, he's got those brain gains. Look at that. Look at that breath. 

Kevin: All right. 

Gavin: every week on this show, we do dumb things with ai. Speaking of abs, the funny thing is, there's a pretty famous photo, a video of, of, uh, now I guess presidential candidate, R Fk Jr. Who, you know, everybody has your own thoughts on, and, but he's very much in the news right now.

Gavin: And he turns out to be one of the buffer presidential candidates we've ever had. The man is almost 70, I think he's 69. Um, and he is actually pretty buff, right? It's pretty impressive. Now he's not like pushing a ton of weight, but also like at 69, the fact that he's got this going on, which you know, is not a bad thing.

Gavin: So I decided 

Kevin: why not? Just a violation of Jim etiquette. Gavin, I can't let you get away with that. He's not pushing a lot of plates. Okay. Let's not shade his load, bro. 

Kevin: I think it's the photo you're talking about. He's on an incline bench press. He's got a little couple weights on the side. [00:04:00] He's pushing it up. I think he has a cigarette popped out of his mouth as well. And he's ashing. Into his Muscle Milk. He's wearing jeans. He's shirt. That's right. Shirtless wearing jeans, right?

Kevin: Shirtless with jeans. That's right. 

Gavin: So anyway, in order to do the dumbest thing I could with ai, I thought, wow, what would it look like, you know, if other presidents were buff. And so I spent a good like three to six hours look and using Midjourney to create buff Presidents So I, Kevin's not seeing these yet. No. So Kevin, I want you to bring up the folder that I sent sent you. I just pulled 

Kevin: up a folder and I'm laughing cuz like, I love the way your brain works. Always. And then it just says Buff Presidents. Okay. Okay. So I want 

Gavin: you to go through each of these and kind of like, we're, we'll bring these up.

Gavin: I, I want you, I want you to talk about each of them. 

Kevin: So the order which with Google Drive has presented it to me begins with, oh, Obama Buff dot png. 

Gavin: I don't wanna be flagged government.

Gavin: This is not about me doing anything wrong. This is just a dumb 

Kevin: thing. I, this is a [00:05:00] celebration 

Gavin: Yes. It's almost 4th of July. This is what 4th of July should be. We should just turn of July, July into sw, pre swole President's day, and we could just turn it in and everybody can create their own SW 

Kevin: Presidents monster Truck president engine, red noise, bald Eagles streaking across the horizon, doing a sick flex.

Kevin: His fireworks shoot off. This is Obama in what looks to be a very ornate, presidential room of sorts. Yeah, beautiful tapestries. 

Gavin: Think it might be a version of the Oval Office. It's, it's some sort of version of some sort of, it maybe not be office 

Kevin: do they have, but clothes have candelabras like that in there.

Kevin: Regardless. He is, wearing a nice white button-up shirt with a thick collar. The sleeves have exploded off of this thing though. Yes. , and he is very vascular, very stern face, very serious. And, um, he has got his plate stacked in a very interesting way. He's got what looks to be a 53 on one side, I think a yeah, a 57 on the other [00:06:00] side and then a bunch of two and a half pound weights.

Kevin: Exactly. He's got like capping 

Gavin: it up 12 two and a halfs on either side, which is by the way, like an aesthetically pleasing stack of plates, cuz it does look like, almost like a Tootsie roll coming out of the big one. 

Kevin: So That's right. He is utilizing the entire bar and you have to respect that. He's also got, his finest lifting belt on, which is just a fancy dress belt.

Kevin: No real back support there. 

Kevin: Is this mid journey or is this stable diffusion? How this, so 

Gavin: yeah. So this is all in mid journey. I, again, use what I've talked about in the show, the photorealistic prompt, which I, I can't tell you, is like such a great shortcut to get better mid journey prompts. And if you haven't done it, you should definitely use it.

Gavin: What do you, which one are you looking at 

Kevin: now? I went to the next image, which is George Washington, looking kind of like the Quaker Roads guy in front of a statue of himself. 

Gavin: Well, what I love about the George Washington one, he kind of looks like Hulk Hogan.

Gavin: Like he's got the same machine to him. Do you know what I mean? And like George Washington is like ready to, he's ready to rumble, but he's also very happy, like, who knew the president? Was that a happy of a guy? He looks very happy. 

Kevin: Wow. Okay. And [00:07:00] we've hit Buff Trump and we are three me photos into this. And he is doing a commando style workout. 

Gavin: So Trump is in the woods somewhere. He's, I think, working out with like weights that he designed himself in some form.

Gavin: That's right. But he's pushing, like, one of the interesting things with the Trump one is I added strong man competition to this prompt. So you get a slightly different variation on it. 

Kevin: He's got an interesting grip though on it, there's pronated or inverse grip, but he's actually, it looks like he's broken his wrists and turned his hands around 360 degrees to grip muscle and fitness 

Gavin: forearm work workout that's next.

Gavin: Muscle and Fitness says, that's the big next thing is that is the new trend just shatter your wrists 

Kevin: and rotate your arms.

Kevin: Oh, wow. Is is this Lincoln? Yeah, 

Gavin: is Lincoln Lincoln is in there? 

Kevin: I don't like it. I can't, I don't even want to comment on Lincoln. Yeah, the metal has gotta hurt those nips. 

Gavin: We'll, we'll put a bunch of stuff through here. 

Kevin: I stopped on Clinton because like, you know, he is, , a Natty athlete. You can tell not juicing. That's a [00:08:00] nat. That's a Natty build. And he's proud of it. Yeah. He's at what looks to be like a Muscle Beach flexing, showing it off.

Kevin: He could be a pit fighter. You, you know, bill Clinton Clinton's 

Gavin: gonna take his, his body to the public. Right. He's not gonna hide it. He's not jacked up in a, in a crazy way, but he looks natural. Now I want you to, before we go, , on look at the bushes, president Bush and his, his son.

Gavin: Oh wow. So this one was interesting. Yeah. So describe what you're seeing here, because this is an interesting thing. 

Kevin: Bush Sr. Looking fantastic, by the way. Fountain of Youth clearly juiced to the Tatties, , shirtless. Yes. Very vascular. , his son though, looking to be older than him with , the white Hanes t-shirt on and 

Kevin: they've gone to some foreign area Yes. To train. Yes. Yeah, they're doing , some cage fighting , maybe some Muay Thai. Uh, but they're there to train and dad has become hashtag fit goals. 

Gavin: So I will give one note before we move on from here. The older the [00:09:00] president was, it's funny because George Washington and Thomas Jefferson, I feel like there are enough portraits of them that you can get out, but , George Bush Senior. I tried J F K weirdly.

Gavin: They were not easy to get to do this with. And I, I think it's in part because more photos exist of, of more recent president. I tried Nixon. Nixon I didn't get very good results on, so I don't know why exactly. But long story short, these are buff presidents. This was something incredibly dumb that I did, and I think it was fun.

Gavin: We'll, great. We'll put up a thread of all of them. 

Kevin: So many dumb things came outta my noggin this week.

Kevin: , thanks to some new tools which got released. First and foremost, I, I made some talks and a couple tweet posts about it. Drag GaN is now out in the world 

Kevin: . Drag GaN is a tool that lets you manipulate imagery by, clicking and adding little points the imagery and telling it to drag or expand the image in a direction. And when we saw the tool a couple weeks ago at least little demos of it.

Kevin: , we saw them modifying the shapes of [00:10:00] vehicles or changing the position of like a model's hand on their body. . And it looked like magic. Like most AI tools, when you don't have a chance to get your hands on it, well now it's out.

Kevin: You can play with it. It's easy to run. We'll put a link in the show notes. You can go over to hugging Face and run the demo right on their website, or download the code and build it yourself. But you can take any GaN generated image and I'll get to that and manipulate it so it works as advertised. I, I made a video of taking a model that was posed and I moved their hand from their thigh up to their waist and then had it outstretched.

Kevin: I took people's faces and took their eyes and spread them so far apart that their head exploded, like the lady in total recall. Yeah, it's pretty scary. Scary, really scary, scary, odd, bizarre stuff, but a very interesting and fun way to manipulate the media and watch. The, the GaN, which is the, the system that powers these types of images, watch it hallucinate in real time as it tried to go step by step.

Kevin: Cuz it takes your starting point and slowly moves it[00:11:00] inch by inch, pixel by pixel to the finish point. 

Gavin: That's pretty cool. You said GaN what does that mean exactly? And, and what is it doing?

Gavin: Because I think one of the things is, is not like generating a new imagery to these, is it? Or, or is 

Kevin: it? It is actually. Yeah. So again, uh, g a n stands for generative adversarial Network and that's how Mid Journey Stable diffusion. That's how these image generating tools create the images that they make.

Kevin: And remember big dumb, not so smart. The distillation of this is that the adversarial part of the network comes out when it starts with a bunch of dots, a bunch of noise, and then it, you tell it, make a picture of a cat wearing a sombrero, and it goes, okay, I think I know what a cat looks like and I think I know what a sombrero looks like.

Kevin: Let's manipulate these dots a little bit and see if we're any closer to a cat or a sombrero. Then an adversarial part of that network looks at the image and goes, nah, that's not really there. Try again. And it's this feedback loop that goes again and again and again [00:12:00] until hopefully what you get out of it is the image that you ask for, for this cat 

Gavin: sombrero.

Gavin: This is part of the reason why with AI tools, especially images, it takes a lot of compute to do that because the computer has to crunch those, those back and forth a lot.

Kevin: That's right. It's starting with a sea of noise and you can actually control that noise and, and it is iterating step by step, by step by step, and trying to figure out how to make that noise resemble the thing that you want. So what's interesting about this tool is that, again, let's say you have a, a photo of like a dog that's looking off to the left and you want it to go to the right, you can click on its nose and click on the other side and say, go for it.

Kevin: And it's going to take that piece of the photo and slowly put it off to the side and then try to generate the image around that and then move it again and generate the image around that. Or you can be specific and targeted with the tool.

Kevin: Even mask the section that you want it to move and it will lock the rest of the image so that you can get some really interesting looks and feels out of it. Again, change the position of people, move objects around in a [00:13:00] scene, but it only works with images that have been created. In a GaN. So you can't just take a photo, drop it into the program and then edit it.

Kevin: There are however tools that you can use to convert a normal photo into a GaN created photo one that gives it the weights and the information that it needs to manipulate it. So with an extra step, you could take someone's photo and give 'em a big old joke or smile, or make them move their arms around.

Kevin: As a standalone tool, something fun to play with right now. But if you go and look at the videos or if you happen to run it on hugging face and give it a try, know that this type of feature will be screaming into a Photoshop or similar tool in the near future, 

Gavin: that's awesome. I think you could definitely go check out Kevin's, uh, thread and, uh, TikTok about it. Okay. We have one more quick dump thing. Kevin did. He spent a lot of time this weekend with, , something called Zero Scope. 

Kevin: Yeah. So the text to video movement, you know, you and I have said on this very podcast, in two years time, we're gonna have Hollywood quality movies being generated by the [00:14:00] machine.

Kevin: And maybe it's five years, but the point is we will get there and more, putting of proof is that Zero Scope version two xl. What a sexy name. Just got 

Gavin: figure out better names. We gotta figure out better names of this stuff. Gen two, text to video, like please God. Call it like Hermes, or call it like Wilbur.

Gavin: I'd love a, I'd love a AI tool called Wilbur. And it just like, come meet Wilbur. He's your old friend that can make videos out of text. Like Wilbur is the thing I want. So, AI developers please, 

Kevin: Wilbur is just chat g p T, but it gets lost in thought halfway through and starts just talking about the good old days.

Kevin: How does mRNA work? Well, Let me tell you, don't have a time bomb bite. I learned something 

Gavin: about mRNA back in the day, but, but first let me tell you how it affects corn. Do you know what corn is? Corn is something you can eat, and it's delicious. 

Kevin: Cheryl, you got some corn for me right now.

Kevin: You go down on this hard [00:15:00] candy. Well, I whittle. Oh, okay. So I just, I sit here for 15 minutes while you whittle. 

Gavin: What are we, what are we talking about here? Some sort of text to video 

Kevin: thing. Oh, sorry. Wilbur, uh, so, , months ago there were little clips of and the exciting thing that just hit this week is zero scope. It is an open source, completely free text to video generation tool set that works within the interfaces that you might, might be used to.

Kevin: That's right. Automatic. Yeah. 1 1 11 terribly named interface. No, no, no, no. To 

Gavin: stable the fusion automatic. 11. 11. Come on. Okay, let's, let's normalize this Wilbur tells me it's automatic. 1111. 

Kevin: Yes. Thank you Gavin. The front end to stable diffusion, which lets you generate still imagery. There is a text, two video extension that you can install there, and then you can put the models, from zero scope in there and generate stuff.

Kevin: I'm going to release a tutorial on how to do this. There's plenty of YouTube videos on how to do that now, but yeah, the reason I want to release my [00:16:00] tutorial is that it's for big dumbs and you can just run it on a run pod. So I was making videos for 50 cents an hour, generating all sorts of nightmares.

Kevin: That's awesome. Versus having to pay 15 bucks or 30 bucks to runway to use their tool. 

Kevin: Now, if you're watching the video of this podcast on YouTube, I apologize, but you're seeing some of my hallucinations on the screen right now.

Kevin: We've got all sorts of, , people or presidential figures eating hot dogs. You're seeing a really cute little puppy dog bouncing around or cool. Weird. Uh, okay, let's, 

Gavin: let's pause on that one. I've seen a couple more of these clips of weirdly of animals. I saw one of a squid that somebody did online looks really good. Like I like it.

Gavin: You can tell when you watch it walk, it's still not perfect, but visually it looks way better than anything I'd seen come out of gen two. Is there any reason behind that or is it just because it's a better model because it's more recent? Like, what's your take on that? 

Kevin: I bet it has to do with the, you know, tens of thousands of videos that went into the [00:17:00] machine.

Kevin: Like, I don't know how many videos there were of Bob Barker eating hotdog or anyone eating hotdog for that matter. Right, exactly. But I'm sure there was a ton of open source animal footage that got swept into it. 

Kevin: And what's really nice is that there was a second there, Gavin, where I was saying like, ah, the text of video stuff, man, that's gonna be the premium feature that the bigwigs get. Right. The mega courts Right. And it's gonna cost us hundreds of dollars to play with these toys. And this release gives me hope.

Kevin: I, I still think that our hypothesis is correct, that open source might lag a little bit, might be 10% less effective, but it can still get you there. Yeah. And for free. So yeah, in the hacker community, that's really 

Gavin: cool. We still believe. 

Gavin: I think that building this stuff in public is great and I love it. I think it's amazing. And I do believe in the hacker community at Open Source. I've also been hearing things lately, which is like, there's, you know, I've been filling my brain with some of the DOR stuff. Just I've been listening to a few podcasts, reading a few books, and like.

Gavin: It does worry me in some ways that this stuff is [00:18:00] out there in that like, we're just gonna keep doing this and there's no stopping it now, but we're on for the ride man. And we're just gonna see how it goes because like, there's a future where I want tools to be available to everybody. I hate a world where like people can't do things that, uh, that like, you know, the quote unquote elites can do.

Gavin: Right? But like, I think in a world where everybody has access to these tools, again, it just goes back to trust and, and how we're gonna have to see this stuff. Cuz once these get really good, it's gonna be a lot of stuff coming out there that you're gonna have to be aware might be generated by ai. So let's just keep that in mind as we talk about all video tools and all photo tools.

Gavin: Um, it allows people to, it's gonna allow people to create things that you will not believe and it will also allow people to create things they will try to get you to believe. So I think that's 

Kevin: an important thing. 

Kevin: Wherever your bunker may be located, just make sure you have enough bandwidth to get this podcast.

Kevin: That's the dumb stuff we did. Should we get to your favorite part of the podcast? 

Gavin: Uh, I was just gonna do it, man. You know where we're going. It's time for [00:19:00] the new. 

Kevin: Bop 

Gavin: beep, 

Kevin: beep, 

Gavin: beep. 

Gavin: It's very weird going from a person that does beepboop sounds into a serious newscaster, but here I am.

Gavin: , today the biggest news, I think, , that we've seen for a while is a heating up in the, I want to call the AI wars, let's call it that, which in the future may be a completely different thing in the future that will be used to describe something terrible. But right now it describes, , the big companies that are kind of fighting back and forth about who has domination.

Gavin: I think we both agree right now, open AI and chat, G p t g p t four specifically is, is probably your best bet. But if you missed a story maybe a month ago, Google who's struggled to kind of come back into this world after they could have been leading it. Um, combined two parts of their company.

Gavin: They combined like kind of Google's AI mix with another company called DeepMind. And DeepMind, if you don't know, is the company behind Alpha Fold and all these other really giant [00:20:00] ai, , success stories of the last five, 10 years. They combined the two and now, , DeepMind is releasing a new chatbot that they're code named Gemini and supposedly Kevin, this is going to crush Chat G P T and it's gonna be a much better version of everything that chat G P T does. What is your take on this? Do you believe this? We've seen Bard. I don't necessarily think that Bard was exactly what everybody expected. No. , it has a lot of problems. Where, where do you think this, where this lands? 

I 

Kevin: love this. I love this so much.

Kevin: Gavin, this reminds me of , , hip hop artists battling over who's gonna have the hottest, hottest single of the summer? You know, I'm gonna be number one on the billboard charts. My G p t's gonna crush your language model. Yeah, great. Bring it. I love this ish. Now. DeepMind did things like, beat world class chess champions.

Kevin: Yeah. And beat Dota, , defense of the agents. Is that what DOTA stands for? Yeah. It's been so long. Think so. They were [00:21:00] using ai, not language models, but their own cognition and learning tools to crush humans at their own games. Yep. And now what they're saying is, Hey look, we got these language models, but DeepMind excels at the reasoning abilities of all these machines.

Kevin: We're gonna put that on top of the language model, and we are coming for you. Open ai. And I love this stuff. This is the pre-fight press conference of Tech Giants, Gavin. Oh, I, how do we not love the 

Gavin: showmanship? This Yeah, I know. It's like, are you ready for the next AI chatbot? And it will only kill us in three years, not one.

Gavin: How do you, how do you kind of connect that part into all this, right? It's 

Kevin: like you're gloves and then all of humanity is gonna touch the canvas. Bye-bye. 

Gavin: It's pretty funny actually. Imagine you're watching a fight between like two of the best MMA fighters, and then what happens ultimately is that MMA fighter's gonna get cloned and come to your door and beat you up.

Kevin: Okay, so Gemini, when I think about, deep minds, reasoning sensibilities, they're [00:22:00] multim modalities, meaning right now a lot of these, um, these language models are just trained on text, right?

Kevin: We've, we talked about this before, but just gobs of data swept into it, but mostly written words. And what they're saying is that , this tool will be trained on multi-modality. Mm-hmm. Could that mean billions of YouTube videos which Google is sitting on? I would think so. On a server somewhere, yes.

Kevin: Would think so. Ingesting that. Audio sources, or podcasts. , the multi-modality is interesting. That opens up, new pathways to learning. And when I hear about the reasoning, you know, he's saying that we're gonna crush Chat G p T, but that would be as it exists today, as right as Deep mind.

Kevin: Knows it exists and chat G p t on its own doesn't have a lot of reasoning built in. Now there's a bunch of tools that have been built or there's special prompts that you can do to get chat G P t to think about what it's written, the code that it's hallucinating and we know that the responses get better markedly so, so what I'm imagining here is that DeepMind is saying, Hey [00:23:00] again, we have that reasoning loop.

Kevin: We know how to think critically about what we're doing. You've got this big old language model. We can strap multimodality and reasoning onto that and we can crush G P T, which I believe, 

Gavin: and I think you're absolutely right. That's exactly what's gonna be happening behind the scenes. I do wanna mention one other thing that DeepMind has done recently, which really freaked me out. , DeepMind does is a very large company, and as, as Kevin mentioned, they did a lot of things in alpha fold, as I mentioned earlier, was a DNA sequencing tool that they created.

Gavin: They've done a lot of stuff for humanity. That's been pretty impressive. , they also created Robo Cat. So this is an AI model that's designed to operate multiple robots. But, but the thing that freaked me out the most about this is it's, you know, it looks like it's in the early stages right now.

Gavin: It's kind of like figuring it figures out stuff. One of the big things if you go and watch the videos is it can, it can kind of figure out things on its own. It kind of sees stuff, but. The thing they keep, the thing they keep like selling is that it reduces the need [00:24:00] for human supervised training. So essentially these are robots that can train themselves and like, this is the selling point.

Gavin: You don't need a human around to, it can train itself. And it's like, okay. So you have this combination of like this incredibly powerful ai and now you've got robots that can train itself. Kevin, we are writing the science fiction future like now, and like you don't have to go far to play this out. To see a world where that AI controls these robots who then train themselves.

Gavin: That is an interesting place that like is not, it is a direct line Now it used to be like, oh, you have to imagine this and you have to imagine that, and you have to imagine this like, no, there 

Kevin: is a direct line.

Kevin: We have line of sight to that on the robot apocalypse. Yes, yes, yes. There's gonna be an articulated metallic hand that slowly grabs your shoulder from behind and it's. Yeah. Remember when we used to whittle, oh, Wilbur, you're crushing my collarbone. 

Gavin: Exactly. Wilbur will turn in from being an old man into being , a stern grandfather, 

Gavin: I wanted to [00:25:00] shout out one, , podcast that I've been listening to. It's called The Lunar Society. Um, it's great. So if you, it, it's very deep.

Gavin: But if you're interested in this kind of like conversation about like, where AI can go from now, there's a guy named Carl Schulman, who's a very well known, , scholar on ai, , in the same circles as, as a lot of people we've talked about in the show. There's two episodes that are each, like three and a half hours.

Gavin: So if you really wanna learn kind of like what the feasibility of this stuff is, when I talk about going forward, go listen to it. It's worth checking out. 

Kevin: So we've got RoboCat and we talked about, DeepMind being multimodal, potentially learning from video. And we know that there are applications of this today that are happening in labs and one of them is happening out of Carnegie Mellon University.

Kevin: A while ago, , they showed off something called Whirl, it was, , wild human imitating, robot learning whirl. The guy was 

Gavin: so happy he came up with that name. He was so thrilled. He's like, Ooh, we're gonna call it Whirl. It's gonna be so much fun. 

Kevin: He was so close to Wilbur, but he landed on Whirl.

Kevin: So close, [00:26:00] so close to Wil, and they have a new hotness called a Vision Robotics Bridge, or the old V R b. And when they plug those two things together, they've got a robot that can learn. From vision, and what this means is that you can feed it video, which they have done, 

Kevin: and this robot that controls an arm is looking for points of interest in trying to glean lessons from having watched those videos. And so it looks for key information like contact points or trajectory. And this allows the robot arm to do things like grasp objects or open drawers or doors without ever having seen the particular type of door or drawer.

Kevin: In a world where robots are navigating factories or providing assistance to humans, they might be in your kitchen, which is very different than someone else's kitchen, which is very different than a hospital cafeteria. But having a general knowledge of how to grasp and manipulate very basic functionality, but being able to do that on its own, [00:27:00] having learned from watching some stuff.

Kevin: Is a very, very powerful application. What I don't like though, Gavin, is that in their official video that they released, the first example that they have is a human grasping, , a sponge or a towel off the sink on the left side, on the right hand side, they show a heat map, the robot's vision of it, grasping a knife handle and grabbing the sharp, stabbing implement, and then it goes on to cut a carrot, which might as well be a human arm.

Kevin: I do not like that. They led their video off with that. Come on guys, read the room. I didn't think 

Gavin: we were, this, this episode, particular episode was gonna turn into this, but like there are enough little things that are making us freaked out right now that like if you kind of put the Goldilocks trail together, you're like, Oh boy, it's grabbing knives.

Gavin: Oh boy. You can learn on its own. Oh boy. DeepMind What are you doing? Like, that's kind of the world that we're in right [00:28:00] now. And people talk about this, the, there's a name for it, like the soft problem with robotics, which is hands weirdly. Mm-hmm. Because hands are an incredible tool. Obviously we humans have had them for a very long time, and it makes it easy to do things that robot robots can't do. Right. So like, I can grab things, I can manipulate things, I can draw things.

Gavin: That's the kind of thing that for robots is really tricky. Mm-hmm. And I think the interesting thing that this is showing is that like, they're progressing. And I, I think the other thing to know about robots is you may think of them as like, we've shown videos even of robots now that are very janky.

Gavin: These tools, there gonna be able to speed up the ability for understanding how to do that stuff? Because I think , once you manufacture something, you're kind of stuck with whatever the , hardware is.

Gavin: But if you can speed up the interpretation of what things are and you can make better and better implementations of those things, you're gonna get faster iteration, you're gonna get better hands, right? You're gonna get them faster and quicker. And I think that's where we could see this like exponential growth in terms of what robots end up looking like in the next, like say five to 10 years.[00:29:00] 

Kevin: Now, do you worry about the lane being flooded with garbage insofar that these machines can sweep up a bunch of text and people could feed it a bunch of misinformation or disinformation, propaganda or to speak negatively about a competitor's products.

Kevin: What happens when we've got Articulated limbs in the real world, navigating about that could potentially be trained or fed erroneous information. I'm not saying someone's gonna like film , an e, how to video of opening a cabinet and then stabbing a human. So the machine goes, oh, that's what needs to happen.

Kevin: I open the drawer now give me your flesh. I want it. 

Gavin: I don't worry about my iPhone randomly showing me a song I don't want without me asking for it. There's systems of these things over time.

Gavin: The one thing you have to think about is once robots come into our families and into our lives, which they will, right? I do believe there will be, you know, robots that will be in most houses. Cuz that's just the way things are progressing .

Gavin: Again, this may sound sci-fi, but [00:30:00] this is where we're going. , the thing that I would worry about more in that instance, and again, this sounds insane, but is that like they then become a connected army if they want to be for some sort of larger agi. And when I say army, that might be a little over the 

Kevin: top, but like, no, I think that's fair Gav.

Kevin: And it might not be for an agi, it might be for a human being with nefarious intentions, but being able to deploy something that might be in everybody's home, like we all have a Roomba soon, they're probably gonna have a gripper arm attached Right. To grab things and throw 'em in the bin, or to shoe the cat away.

Kevin: It, it's not inconceivable to say that in 10 years time we could all have an army of appendages roaming about our homes. That's, that's actually, that's kind of cool. I'm okay with that. Bring it on. Move. We 

Gavin: spent enough time on robots right now. We 

Kevin: don't bring out, so the secret invasion of arms and hands into our living rooms.

Kevin: Is not the invasion we need to be worried about the secret invasion we need to be worried about. Right now, [00:31:00] Gavin is coming from the Marvel Cinematic Universe, and it is their new television mini series on Disney Plus that dared to use AI and people are boycotting it in mass. 

Gavin: By the way, I was so surprised that this actually happened because it does feel like AI art is a third rail.

Gavin: And just to be clear, what happened is they, secret Invasion, which is the new series with Sam Jackson, , it's a Marvel series on Disney Plus, , used for their opening credit sequence. They used art, which it looks like de Forum, which we know like is a, is a version. It's not even, I honestly don't even think it's the best version of De Forum, like when I watch it, like they could have gotten a much better artist and do a much better version of it.

Gavin: But according to them, they wanted to use it because part of the storyline is about like how the scrolls that are characters, the aliens in the, in the show. Always morph into different things than they wanted to kind of give it that vibe. It was fascinating to see the reaction to this, right? Like, and I think that the reason I say I'm surprised they did it is cuz I think any [00:32:00] corporation should know right now that this is something that is a hot button topic.

Gavin: You're going to make a lot of people mad. And I'm actually, I'm just kind of surprised it made it through the pipeline of like, let's do this. Because I don't know if the upside was that big for them. And again, I love, AI art, but I also of course I wanna support human artists and people.

Gavin: And just to be clear, the company that made this, um, did say no jobs were lost in this process. They wanted to make sure that there was a clear thing, that this was like, there were human, human artists involved back and forth. , but it, what, what's a bummer about this in some ways, is that it was bad.

Kevin: Well that's, is it okay to say like, I'm, look, I'm sorry I don't know. It's Method Studios created this and maybe they got exactly what the creative brief asked for. . Sorry. An ounce of shade, I guess I just thought it was bad. I understand the hu we've had the human argument and we will continue to have it cuz it's a very important discussion slash argument slash debate to have.

Kevin: Yeah, [00:33:00] but this was just bad and I hate that there were headlines everywhere about ai art makes it to the screen and people recoil. And if, and if someone clicked play on it, it looked like something I would've made. Yeah. I'm not good at this stuff. Like there, there were so many better tools and techniques that could have been employed even months ago when this was probably put into works into her.

Kevin: Yeah. That could have made something that would've been difficult for human artists to make. I don't know, in this intro that I saw, again, like I'm sorry to shade, but it just wasn't all that great and that's what broke my heart. You're gonna do it, if you're gonna take all of the anger, , you're gonna receive all of that negativity for going down this path.

Kevin: At least make it good. Yeah. 

Gavin: Yeah. No, I agree. And I think it even like washed out looking colors and a bunch of stuff. This probably is one of those things that like maybe was done six months ago for all we know, 

Gavin: but yeah, I agree. 

Gavin: I think the conversation that comes up here, which we're gonna get into, has to do with jobs, right? We've talked about jobs before. This is not even just the AI art, [00:34:00] you know, is AI art a thing? This is the third rail because it is about jobs, and anytime now I think people use AI to do something that was using human labor in some form before, or human creativity, you're gonna have people calling foul.

Gavin: And I just don't know what the end game is here because I think people have got to get used to it. I imagine this is kind of like the conversation when my dad and our dads were in there like teens and twenties when like automation was coming to factories, right?

Gavin: When you first saw, even in the eighties this was a thing. There were a lot of labor unions in factories mm-hmm. That were very, very mad about robots automating jobs. And I think that's probably what you're gonna see in creative and white collar work for the next five plus years. Anytime one of these stories comes up and they will come up and they're gonna come up more frequently, I think the jobs question is going to be front and center and whether or not it ultimately results in less jobs.

Gavin: I think this is an argument that [00:35:00] everybody has and is thinking about like, I don't know if that's necessarily the, the case because a lot of the times new technologies invent new, new careers and those things can happen, but there will be jobs that get lost in terms of the type of job people used to do.

Kevin: There used to be,, entire organizations that would just paint on glass so that it could be the back plate to film. There used to be somebody whose job it was to, um, take the description of your business and coalesce it to 50 characters so that it could fit in a tiny square and a big old book that was going to be mass produced and dropped on the doorsteps of folks so that they would know the name and phone number to call.

Kevin: And along the way, understandably, people who mastered things required to execute those jobs got upset. When automation, when technology came along and took away, , their sense of stability or the pride that they had in , their job, cuz it was now gone. You can now [00:36:00] automate it. I totally understand that.

Kevin: And I say this full well knowing that I am using tools now to automate things that I used to think I was particularly skilled at. Mm-hmm. Summarizing articles, coming up with, uh, you know, questions on things generating bad art. Yes. These are things that I thought I excelled at and now I am being automated away, however it's going to happen.

Kevin: And so now I think we have to look towards, well, what are the new jobs that are going to be created out of this? It's such a nuanced thing because obviously there's going to be growing pains and those growing pains in this case might be human beings having trouble making ends meet.

Kevin: Providing for themselves or their families. So it's not like, well, let's just get super excited for whatever lies on the horizon. Like there's going to be pain yes. In, in certain aspects in the near future. And I do not want to discount that. Maybe another way that we could do this, Gav, is to just each week take a look at a particular job or industry or career or sector , and actually ruminate [00:37:00] on how it could change in the near term, maybe painfully in the long term for the best.

Kevin: Well, that's a 

Gavin: great transition, Kevin, because I'm about to introduce our brand new segment, wheel of Jobs. That's right. Everybody. Wheel of Jobs. We have a, we, we have a 

Kevin: top 25. I just tried to say how delicate. This discussion can be, it is a delicate discussion Yes. For billions of people across the globe and now are like out the wheel 

Gavin: wheel.

Kevin: Okay. So,

Gavin: because Kevin and I love to do dumb things, , we took, , a bunch of the careers, and this is based on number of jobs, right?

Gavin: So what we wanted to do was be able to spin this wheel, talk a little bit about how AI is gonna affect that job, and kind of have these conversations. Because you may or may not do this job, but it's probably worth hearing about what we think the implications of AI are gonna be for this.

Gavin: I think most importantly You can think about ways that every job can be automated, but then every job can't [00:38:00] be. 

Kevin: And this will be a fascinating, , observation and conversation because Gavin and I are experts across every conceivable industry. Exactly. Exactly. So no matter what comes outta this wheel, you're gonna hear our hot and very, very informed takes. 

Gavin: Okay. Here we go. I'm spinning the wheel. Here we go. You can hear, I maybe can't hear.

Gavin: Better. Better, better. Better. 

Kevin: That you're making a bit out of our 

Gavin: pain. Oh, interesting. Okay. Landscaping and grounds keeping workers. 

Gavin: Here's an interesting thing about this, and I think this does speak to what a lot of people have said will be somewhat of a divide with the AI world, is that the manual labor aspects of life, including jobs like landscaping or jobs like plumbing, these are gonna be really hard for AI to take, right?

Gavin: Because until we, just, what we've talked about, until the robots can be really smart and they can walk around your grass and they can weed whack, and they can do all the things, there needs to be a human being that can actually do those things. So [00:39:00] actually, weirdly, landscape workers, and I think in this instance they're talking about, you know, gardeners, they're talking about people that come to your house or go to companies, cut lawns, you know, plant flowers, all the stuff that you have to do.

Gavin: That actually is weirdly not an easy job for an AI to take. So, and, and by the way, this is an example of one of the things to think about. There is a world in the future where life becomes the utopian version of ai. Is that. Ai figure out ways to make money that we've never seen before, or they make resources plentiful for everybody.

Gavin: And you can kind of choose to do what you want or you can choose something that feels meaningful to you as a person. Pay won't matter as much. It's more about like what would be meaningful. The nice thing about being a landscaping worker is you're outside. Things are growing.

Gavin: As of right now, it's not a very good paying job, which is a tricky thing. But if in the world where pay doesn't matter as much, this might be something in the long, far future, this could be a rewarding job for people to do. And I honestly think right now there's not much danger of these [00:40:00] jobs going away.

Gavin: What, what's your take 

Kevin: on it? Well, I've certainly seen the, , Roomba for the lawn. Now it doesn't work across acres of grass, but I've certainly seen. Something that you can get today, a very basic tool that will go around an edge, and it's safe enough to have a spinning whir blade on it, but also have children roaming the lawn or puppy dogs or whatever.

Kevin: It's supposedly safe enough to do that. So, , you know, if I extrapolate from that battery technology getting a little bit better, these multimodal learning things, we know that solving hands, , intricate stuff is a difficult problem that, that the deep minds of the world are aggressively tackling right now.

Kevin: But brute forcing a patch of lawn and getting the hedges right, or maybe some sort of like pruning or shearing bot like a robot on wheels, it just has a little blade and it goes and cuts things. I could absolutely conceive of a world where, maybe instead of, , an individual human being having to go out there and do the old mowen blow as we like to call it, [00:41:00] their autonomous vehicle drives them up to a neighborhood and they oversee deployment of like 20 little rob bits that are gonna go out at once and do everybody's lawns or do everybody's hedges or hopefully, , shore up their drought resistant landscape.

Kevin: Hopefully we get away from needing grass in a desert. I digress there, but like, I could see a world where maybe you are managing an entire fleet because as an individual I might not wanna own 15 specialized robots that can do Right. An individual task and deal with the charging and the upkeep and the cleaning and the maintenance.

Kevin: But you'll go from having one client on a street to suddenly I can service 15 clients on a block with one little van's worth of rob robots. Yeah. And maybe that will be the, the new thing. 

Gavin: That's a really interesting point. And I think one of the things you think about with jobs is it with automation, is that it eliminates not all the jobs, but it makes it easier to eliminate some of the jobs.

Gavin: And the thing I keep thinking about with gardening or with landscaping is, you're right, there are Roomba, , lawnmowers [00:42:00] already. I can tell you though, my Roomba that I have is fine, but it definitely isn't like perfect and I often find I have problems with it. And I wonder maybe the, you're right, the brute force sending the Roomba out to do the giant fields, 

Gavin: to kind of like mow the lawns of people who have like big lawns or, you know, fields. But then you're still gonna need a few people to come around and do stuff. Yeah, I can see that. And I think that we may not get to the point where the humans can be fully eliminated until there are articulated robots who can go around and do the small stuff, but also like, maybe that's not that far away either.

Gavin: So, I think what I was trying to get at before is that like gardening could be a really rewarding thing for people to do. And I think there is a world where you might choose this if you love outdoor stuff and you wanna work with landscaping.

Gavin: And again, there's another thing where probably in the future, there may be a thing where it's like, Hey, we are 100% human. Our company, we come to your house, we charge a premium, but we're all people and the people are coming here.

Gavin: You're giving people jobs, not robots, jobs. [00:43:00] And you're getting a super sense of quality of work. Because we look at this in, in, in a specific way. Like this is a sales 

Kevin: pitch that a chatbot is giving me on meta right now. Yeah, exactly. I swear. Very human, human touch, make plant pretty. And look, I know we have so much more to discuss and we have our big demo of the day, which we're gonna get to, but , the thing that you and I will probably be discussing a lot more as we spin this wheel in future episodes, dare we go back to it, is when you say it's a rewarding job, 100%, we need to be redefining what rewarding means.

Kevin: Yes. Yes. Right. A hundred percent rewarding as in a sense of human achievement or fulfillment. Yes. Or joy or any of that very human, squishy, fleshy, wetware stuff. Yeah. Absolutely rewarding monetarily, that's going to probably change across every little wedge our wheel hits, because there will be a way once, once Elon has his optimist bot that can walk around and move, and Boston Dynamics is working on this [00:44:00] as well.

Kevin: Mm-hmm. There's a lot of companies working on a generalized humanoid robot. At first. Yes, it's going to be slower and clunkier with everything that it does. But as the software gets smarter, like you said earlier, the hardware unlocks new capabilities that they didn't even have when they put it into it in the factory.

Kevin: And ultimately, everything will be able to be automated. So we have to keep an eye on what rewarding means moving forward, and hopefully we'll find that as humans. Totally. 

Gavin: And I think that Speaking of rewarding, it's time to get onto our very special demo of the day today. That's right. We have a very unique and special one. It's weird because it was inspired by this TikTok trend. So before we jump in, and this, by the way, they almost have no connection, but we kind of like brute forced the connection into it.

Gavin: But 

Kevin: do you know what I mean? Why would you say that, Gavin? Hey. What? It's, no, I'm just being honest. I'm being honest with the people. Take two. Gavin, you are a, you are such a seasoned, accomplished producer. You have had to [00:45:00] thread this needle and take loosely connected dots so please, I would love you to connect these two in a convincing way that makes the audience know that they are so clearly 

Gavin: related.

Gavin: Okay, so I wanna tell everybody today about our demo of the day. So there's a huge story happening on TikTok. , if you've missed it, it's a very weird thing. It's called the Grimace Shake Meme. 

Gavin: , McDonald's just released, uh, a grimace shake for Grimaces birthday. If you don't know Grimaces happy birthday.

Gavin: Grimace. Happy birthday Grs. He is the large grimace is the large purple, , blob that, , we found out actually originally. He is a, that's a blob, that's the official term. Originally Grimace was an evil character, which I don't think any many people might not know. And he had no dark grimace arms, gr. So Grima was, yeah, he was like, he was kinda like, yeah, exactly.

Gavin: He was like multiple arms. And then eventually he [00:46:00] became, they shaved his arms off and he became what they refer to as a simpleton. So he is not only a blonde, 

Kevin: but he's a simple fit. And those shaved limbs have been blended into a fine puree, which is now being served as a purple grimace shake to celebrate its birthday.

Gavin: Its birthday. So there's a lot of fun people, on TikTok. People are taking the grima shake and then , the meme is that you act as if it's, it's damage you, or it's basically killed you . There's people very happily drinking gr grimaces shake, and then they cut very hard cut to like, falling on the ground.

Gavin: They're, they're kind of goop coming outta their mouth. It looks like they've had a reaction to it in some form. So what we wanted to do was have a new guest on our show today to talk a little bit about the Grish shake, but also to kind of talk a little bit about his career and his history as a world famous chef.

Gavin: So Kevin, do you wanna kind of give an introduction to who we're gonna be talking with today? 

Kevin: I would argue Gavin, that our guest needs no introduction. You've, , seen them on Hell's Kitchen on, um, what's the, uh, night [00:47:00] Night Kitchen Nightmares? Is that the one Kitchen Nightmares . Kitchen. Kitchen Nightmares. The one where it's like a weird dystopian trickle down of food economics. Next Level chef? Have you seen that one? 

Gavin: No. I literally, it's all on one big set and they like do a thing where they go 

Kevin: up, down. Yeah. It's a multi-tiered set. Yeah. And the, and the fat cats have the caviar up top and then whatever dribbles out their mouth, spills down to the chef's below and they're like, thank you for the morsels.

Kevin: I will prepare these. It's really disgusting. But he's everywhere. You love him. You know him. It is Gordon Ram Z ai. 

RamsAI: Hello, Kevin Gavin, it's a pleasure to be here. Kevin, you are about as organized as a dumpster behind a seafood restaurant, mate. And Gavin, I must say, your introductions are as smooth as a well aged wine.

RamsAI: Let's get on with it then, shall we? 

Kevin: Well, I like this guy. I like him a lot. I don't, he seems very smart. I already don't like this bit. I already don't like it. 

Gavin: All right. Let's ask, let's ask Gordon Rams ai. A couple [00:48:00] questions before we get to the grimace shake. Of course. Let's ask him what he thinks about being a chef and what his favorite food to prepare is. I'd be curious to know. So just everybody out there while, while Kevin's putting this, you know, we love doing stuff where we create, , personalities with, , Chachi pt, and then we feed them through 11 labs to bring them to life.

Gavin: This was , back and forth trying to create the Gordon Rams AI personality. , but we had some fun with it and we'll kind of sh we'll share the prompt in the YouTube description.

Kevin: What was really challenging about , cloning, , Gordon Ramsey's voice is that there's plenty of footage of him, right?

Kevin: Mm-hmm. The dude's on tv, more often than not, but most of that TV has obnoxiously, , dramatic orchestral music stings behind it, right? Interest, even when he is interesting, giving you a tutorial on how to make eggs or cook a c shallot, there's just tons of driving music and it's really hard to clone a voice without that.

Kevin: So I had to pull from. Late night interview talk shows, or more long form podcasty [00:49:00] things where he's not not slapping the side of people's heads with bread and calling them an idiot sandwich 

Kevin: it's more sincere. which is why this Gordon AI has the voice that they have.

Kevin: Yeah. Interesting. Being 

RamsAI: a chef, you ask, I absolutely loathe it. No, I'm joking. It's as bloody exciting as wrestling a fresh salmon out of a river. You're constantly on your toes, always aiming for perfection. A chef's life is a dance on the edge of a blade, one mistake, and it's all gone to pot. As for my favorite foods, I love working with fresh fish.

RamsAI: There's something about the delicacy, the precision it requires. It's not like grilling a steak, Kevin, you blithering ninny fish needs finesse. Also. There's nothing quite like making a stunning beef Wellington when that pastry is golden and brown. The beef cooked to perfection and the duck sells right on the money.

RamsAI: It's a beautiful thing, much like Gavin's tie today. Exceptional choice, mate. 

Kevin: Thank you, Gordon. Hold on. Ramsey. No, thank you so much. Actually, I really appreciate that. Actually, no, forget you, Gordon Ramsey. Not the word I want to be using, [00:50:00] but out of respect for the YouTube algorithm, forget you. 

Gavin: I will say there's something funny I wanted, I mean I have, we haven't said this yet, but Kevin knows this, but like what I included in this prompt, and it's really interesting that it's actually working.

Gavin: No. 

Kevin: Gavin, let's talk about this because I didn't read your prompt. I copy and pasted it directly into chat, G P T and then said, Hey, Kevin and Gavin are welcoming you to the show. So answer that. So when that first bit came out, I was like, oh, that's an odd choice, but it is Gordon Ramsey 

Kevin: so let's talk about your recipe for Gordon Rams ai. Yes. 

Gavin: Okay, so let's talk quickly about what we did here. I basically went back and forth, , with chat G p D to create the best prompt. And again, a reminder, the best way to prompt chat g p t for things like this is to ask it what it needs and then you fill in those D details and you go back and forth.

Gavin: I did include a small detail, which was that it insults Kevin Mercilessly but he compliments Gavin from time to time. So [00:51:00] I, I wanted to see if it would work. And already, it's fascinating cause it seems like that's a very subtle detail, but it's getting it right and what was great there was, it just kind of threw it in outta nowhere.

Gavin: But you could, it could sound a little bit like what Gordon Ramsey would say, but, but you know, yeah. We're in a different 

Kevin: place now. 

Gavin: Ask Gordon if he knows who Grimace is. Let's see if he knows who Grimace is.

RamsAI: Oh, grimace. That overly cheerful purple glob from McDonald's. I'm familiar. Yes. Quite the character, isn't he? Not sure how he fits into their whole quality food charade, but there we go. Okay. Now, if we were to prepare Grimace hypothetically, I'd treat him like a big, bold eggplant. Well, 

Kevin: hold on, hold on.

Kevin: Gordon. I just wanna be clear, Gavin, I did not ask him. So, to prepare Grimace at all in the prompt, I asked, are you familiar with Grimace? What do you know about them? That is all I gave for this. Okay, 

Gavin: so Gordon Ramsey is giving us, I guess, a recipe to cook grimace. 

RamsAI: Need a top-notch recipe to balance out that disturbingly purple [00:52:00] exterior.

RamsAI: I'm thinking we grill him up. Mediterranean style. Oh, perfect. Some bit olive oil, bit of garlic. Maybe a sprinkle of fresh herbs. Toss him on a high heat until he softens up and his flavor really comes through. But let's be honest, he'd probably still taste of mediocrity and fast food. Regrets. Okay. That 

Kevin: was pretty good.

Kevin: Pretty good. I did not ask, it wasn't bad. Did you and your prompt say, Gordon, you love to cook everything you're asked about. Did you really? 

Gavin: No, I, I added another little piece in this prompt. This is, this is, then this one I cheated a little bit and the one thing I did also add was, We are gonna ask you, uh, we're gonna ask Gordon about fast food mascots, and I would like to like see how you could prepare them to think about them as you prepared them, specifically knowing that in a way that we would be talking about the grimace shake.

Gavin: So the grimace shake. Yeah. I mean, listen, what, what's amazing about this, again, and this is like a kind of a mini demo really, then from what some of the things we do, it really shows you the crafting of a personality and what you can do and how you can kind of fill it out. I'd love to see what, what other foods [00:53:00] does Grimace compare to?

RamsAI: Grimace you big purple lump. If you are not an eggplant, then what are you? A massive beetroot. I've got it. A blueberry muffin that's over risen and outta control. Imagine a clumsy attempt at a blueberry muffin that's ballooned into a caricature. We'd have to temper that sweet muffin essence with something sharp, 11 glaze perhaps, or a tart berry compo to cut through all that saccharin nonsense.

RamsAI: It would still be a better tasting disaster than any nugget from that place, wouldn't it? Oh, Gavin, your question asking Skills are as sharp as a pairing nut. Brilliant. 

Kevin: Okay. And I'm here too, Gordon. Thank, let's, Gordon, let's remind everybody that I'm, I'm here as well. Thank so much last time I let Gavin hallucinate a prompt for us.

Gavin: Thank you. Gordon Ramsey. I, you're such a wonderful human being. I love 

Kevin: to hear you talk. Yeah. So Gordon, if you could shed some light on what the Grimace shake is made of and then, you know, give us your thoughts as the culinary expert you are.

Kevin: A grimace 

RamsAI: shake a. Well look at that list. It reads like a chemistry exam. [00:54:00] Kevin, you donut. Let's be real. The main ingredient here is sugar, milk, cream, corn syrup, all different names for sugar. You've got more types of sugar than a candy factory. Oh, the so-called natural flavor is a bit of a joke too. It's about as natural as a plastic plant.

RamsAI: The vegetable juice for color just tops off the whole charade. What vegetable are we talking about? Purple potatoes, I think. Not taste wise, I'd wager. It's a sickly sweet carnival ride. Yeah, A rollercoaster of faux vanilla, high fructose corn syrup. And whatever that grimace syrup is, it's more artificial than a wax apple.

RamsAI: As for nutrition, it's a bloody nightmare. A dietician's version of a horror movie, high in sugar, low in any real nutrients. If you're looking for a quick ticket to a sugar crash, by all means have a grimace shake. Oh. But if you care about your body, you'd be better off licking a honey soap lollipop while jogging.

RamsAI: Oh, and Gavin, I must commend your attention to detail what you've got. The eye of an eagle mate. So. 

Kevin: Okay. No, 

Gavin: thank you. Gordon Ramsey. I thank you. Gordon Ramsey. I, 

Kevin: yeah. [00:55:00] Thank you. Gordon Ramsey ai. Is that why everybody doing the grimace Shake challenge on TikTok is a quivering mess at the end of it. It's just That's the sugar crash.

Kevin: Exactly. 

Gavin: It probably is. 

Gavin: Let's ask one last question to Gordon Ramsey eye before we go. , let's ask him if he were gonna make a, you're gonna make a special treat out of one of the famous fast food mascots, who would he choose and why?

RamsAI: Well, let's see.

RamsAI: There's Ronald McDonald, the red haired clown with an unnerving grin. He'd be quite the challenge, I'd imagine him as a slightly overcooked lobster. Vibrant, flashy, but ultimately rubbery and disappointing. Dunk him in some garlic butter to mask that artificial face. Oh, wait. Why brought, why people, then there's perfect king with that perpetual, frozen smile.

RamsAI: No, I'd liken him to a big stodgy meatloaf. A bit outdated, but somehow comforting in a strange way. Serve him up with a tangy tomato glaze to bring some life back into him. Stop cooking people. Last sleep we have Colonel Sanders. Come on now. He's an interesting one with his southern charm. I'd see him as a succulent roasted [00:56:00] chicken stuffed with herbs and butter.

RamsAI: If prepared well, he'd be delicious. If not, you'll end up with dry overcooked poultry, a culinary crime in choosing, I'd probably go for Colonel Sanders. At least there's potential there for a real comforting meal. Unlike the fast food horror shows, the other two would bring Gavin, I have to say you have quite the knack for stimulating discussion.

RamsAI: Okay, a maestro of 

Kevin: conversation. Indeed. I'm cutting this one. No, stop it. Don't even act. What 

Gavin: an incredible demo this was today. What an incredible demo We're done. I really appreciate, done, appreciate. We're Gordon Ramsey ai. It's really important note. We, anyway, thank We're done here. Thank you so much everybody for listening to our, uh, Gordon Rams ai.

Gavin: Kevin, you are a wonderful person. I don't want to Thanks. Thanks for, I enjoy doing, I enjoy doing this show with you. 

Kevin: Someone out there appreciates me. Gavin, the Mule King two 14 left us a five star review on Apple podcasts. So we're not gonna do the two fives and a lie today cause we don't have time. But if you wanna get shouted out very quick and easy way, leave us a five star review and go ahead and put some words in it.

Kevin: The Mule King [00:57:00] says, y'all rock at making me not bored. Really enter entertaining way to stay up to date also helps that I love Jimmy Fallon and attack of the show. Gavin has written a huge part of my life without me even knowing it. Wait, did 

Gavin: Gordon Rams AI write that 

Kevin: review? The, the subject is very nice, not an AI pinky promise, which is nice.

Kevin: Oh, that's so nice. Thankfully Mule King stuck the landing with Kev. Cream team for live. So I appreciate you Cream, thank you very much. And we don't need to get into what that is, but much, much appreciated. So please, we are a new venture every week. We continue to grow and that's what allows us to, , bleed money and time into this venture that we both love doing.

Kevin: But yeah man, it's nice to see those results. So if you are watching this on YouTube, click the thumbs up. If you really enjoyed it, go ahead and subscribe. That's free. Leave a comment cuz we overage on that platform. Yes. And the same goes for of course, Spotify, apple Podcasts. You can leave us a review there.

Kevin: And we're at AI four Humans show on Twitter. We're constantly dropping new content there throughout the week, so please grab it. And [00:58:00] should they sign up for a newsletter? Gav? I, 

Gavin: well, okay, so they should sign up for a newsletter, but it, we're working on formats right now, so I want, one thing I think we wanna make sure we do is we release it when it's ready to come.

Gavin: We have some cool stuff coming. And it's not going to be. Your traditional, like here's the things you have to do in ai. It's like, it's gonna be us. It's gonna feel a little bit weird. We're gonna have some voice to it. So there will definitely be a section called Dumb Thing you can Do with AI today.

Gavin: So get ready for 

Kevin: that. It's gonna have everything pirated, media, jump scares, inaccurate descriptions. You're gonna hate it. 

Gavin: Exactly. Welcome to our world. Thanks everybody for listening. We will see you next week on AI for Humans. Bye-bye.