Transcript for:
Ranking of AI Tools and Discussion

Model agnostic, ship fast, solve a specific problem. When people are making an AI startup, they should be looking at cursor and perplexity. It's pretty damn impressive. Where would you put that one? S tier. That's one of the few that we can totally agree, like S tier. That's one of our main go-tos. One of the most useful AI tools available out there right now, in my opinion. Hey, welcome back to the Next Wave podcast. I'm Matt Wolfe. I'm here with Nathan Lanz. And today we're going to do a super comprehensive AI tool breakdown. In fact, we're going to do an AI tools tier list where Nathan and I break down kind of all the AI tools that we've used in the past are using right now. The most talked about AI tools, and we're going to put them on a tier list from S to F and really rank where we would put them and also talk a little bit about how we're using each one and give a little context as to. why we're ranking them that way. So you're probably going to listen to this. You're probably going to completely disagree with some of our rankings. You're probably going to be frustrated that we excluded some of those tools. Let us know all that stuff in the comments. We want to hear it. We want to know where you would rank them. We want to know what tools we're missing. And if you like this episode, we'll probably do more like it. So let us know that as well. But without any further ado, let's go ahead and just break down this tool list here. So if you're listening to like the audio version of this podcast, this might be one that, you know, You might appreciate the video version if you actually want to see the tier list. We'll do our best to talk about the tool and tell you where we're putting it on the tier list. But this is sort of a visual episode. I'm going to start all the way on the left. Let's start with Grok here. So this is the XAI logo. I don't think Grok itself has its own specific logo. Maybe it does. But I just pulled in the XAI logo. Where would you put Grok slash XAI? Yeah, I mean, I hate to not give it like a super high rank because like obviously I'm a fan of Elon Musk. And I'm one of the few people that I know who thinks that long term that Grok and XAI is going to be the main competitor to OpenAI. Like I think they're going to be bigger than Anthropic is my current opinion, which seems to be controversial. But right now I would probably put it as a B. It keeps getting better. Like the version two is a lot better than most people realize. It's not as good as ChatGPT or Claude. But it is nice that you can talk to it about pretty much anything you want to. It is less censored. And the image generation is pretty good. The image generation is a lot more free to make whatever crazy meme you want. So I think Grok is like a solid B right now. I think long term it could be an S. Yeah. So for me, just based on like how often I use Grok, I'd probably put it like down in D. Like that's not me saying it's a bad product. Yeah. But. for me it's saying that like all right if i need to go use a chat bot right now to get an answer to something grok isn't going to be my first choice it's not even going to be like my top three right like gpt claude perplexity you're probably gonna be my top three grok would probably fall number four for me for actually going and using it to ask questions right now one thing that i've noticed it's kind of surprised me is i thought it would be a little bit better at getting real-time data from x somehow yeah And I don't think they really nailed that yet. So that is one thing that kind of sucks is like, I went there several times like thinking like, oh, I can ask it some question about an ex-user, about my own profile or whatever for some kind of insights. And it kind of is not that great at it. So I don't know, maybe C, maybe agree on C. I don't think I'm going to keep it in D for myself because I think that like the biggest redeeming factor of Grok for me is that it uses the Flux 1.1 Pro model. I think it's Flux 1.1. But- You know, Flux has three models, right? The AI image generator Flux. It's got Flux Snell. It's got Flux Dev and it's got Flux Pro. Flux Snell, I believe, is like the free version that you can use. Flux Dev is a little bit better model that you can use with like the API. And it's improved a little bit above the Snell model. And then you've got the Flux Pro model, which is actually kind of expensive to use, but it creates really, really good AI images. And... Grok is actually using the Flux Pro model in there. So by using Grok, you actually get access to like one of the better AI image generated models is out there. So for me, I think I'd probably solidly put it as a C just because it's not my main go to, but it's got a great AI image generator in it. I do see a lot of memes and stuff on X that are definitely being generated with Grok. Oh, yeah. Anytime you see like an image that has like blood or like an actual like famous person in it, like a lot of the Kamala and Donald Trump and Elon Musk memes that you've seen, most likely created with flux, most likely done inside of Grok. All right. So next up, let's go chat GPT. Where do you where would you put chat GPT? Yes. Like GPT is S. Like I still use chat GPT more than anything else. So you're going to chat GPT more than Claude, more than perplexity? Yes. Yeah, I still am. At first, some of the I felt like Claude was getting better in terms of like the projects were better than custom GPTs. And it still is slightly, but like OpenAI is already caught up there, too, where like there's more like you can. put multiple files into the custom GPT and things like that. So I use it daily for translating with my wife. I use the voice daily now. So I use it that way. I still have a custom GPT that I use, that I created for helping me write threads on X. I've been using that daily. And also I have a thing that I use for tracking the calories I eat and the protein I intake every day for working out. Yeah, yeah. I use that every day. So like, and like, I'm sure I could do that stuff in cloud too, but I'm just finding like the combination of features that it has. It still has just, it's my main go-to. So I, for me, I, it's not my, it's not my first go-to I'll probably agree and leave it in S tier because I do think that chat GPT is like constantly setting the bar for everything else. Right? Like, I feel like I probably go to cloud and perplexity a little bit more than I go to chat GPT, but You know, they've got their 01 model, which is the best reasoning logic model that exists right now. According to pretty much any sort of benchmark you look at, GPT-4 is kind of leading the pack. You know, almost any time you see a new model come out, you know, Google Gemini or Claude model or Lama model or any model, really, who do they compare it to? They're always like, oh, it's almost as good as GPT-4. It's right up there. on par with the benchmarks of GPT-4. So I almost feel like you got to leave it in S tier just for the reasons that it is like setting the bar that every other company is trying to get to. Yeah, and we're talking specifically about ChatGPT here, but if we were talking about OpenAI as a company, definitely S because it sounds like they have more stuff in the works with the agents. I think they're ahead in agents. And like you said, with the reasoning model, they're definitely ahead there. And... I might not use that every day, but like I am seeing like tweets from like real scientists saying like, I'm using this every day now. And holy crap, if O1 is better, as people are saying, they're super pumped about it. And it's like what one scientist said that it's already O1 preview that a paper he was working on, it helped him get it done in two days, what normally would have taken him a month. And so, yeah, I would definitely put them as tier. Cool. Yeah, I'm not going to I'm not going to fight for that one too hard. It's not my first go to. Yeah. All right. So next up is Claude. For me, I put Claude in S tier because it's probably the one I go to the most. I have tons of custom projects that I've built out in Claude. You know, like I've got a project that helps me write scripts. I've got projects that help me with my notes for making YouTube videos. I've got all of these various projects that I've made. So when I need to make a script, I can literally drag and drop an article in there and it sort of rough drafts the script for me, or I can pull in all of the URLs from like a YouTube video that I'm putting in. And it creates like a list of resource links for me really clean and easy that I use in my description. So like Clod is probably my second most used tool behind perplexity. So for me, that one's S tier. And I also really, really love the artifacts, right? Like I love that I can have it generate code and then it will actually show me what that code looks like over on the right sidebar. That to me is like really, really valuable. So for me, Clod is an S tier. Yeah, I mean, I can agree on S tier. I guess if it was my own tier list, I'd probably put it A. I just, I've always felt weird about the fact that it's like basically an employee from OpenAI who left and tried to do his own thing. And like, and then he's just, and for the first year, it was like basically all copies of Chachibiti. So like, it's like, there's just a weird feeling there. Like, and it's called Claude. I don't know, I hate the name. And also I've noticed recently, like people said it was better at writing. I'm just not seeing that lately. Like I see it's like pretty much the same as Chachibiti. Maybe on code it's better, you know, it seems to be. I think that'll change quickly. Artifacts, definitely give that to them. Like, artifacts is amazing. Like, that's like the one invention that they've created that's like actually really, really cool is artifacts. Yeah, well, they also showed off the computer use recently as well, where you can basically give it instructions and it will actually click around on your screen and stuff. Yeah, which no one seems to be... using that and my understanding from like hearing from friends is like that basically OpenAI already had something like that like for a while and it's just like I think Clyde tried to beat them to it but then like they're also probably behind there and so it's like yeah it's just not useful yet so what do you think do you think it should stay in S tier or do you think it should bump down to A tier because it's And you'll stay ahead of them with competitor analysis tools. If you want it, you can grab it at the link in the description below. Now back to the show. I mean, I do think it's ATR, really. And the other thing, too, I've noticed is people used to say that, you know, ChatGPT was super woke and it was super left-wing. And maybe just slightly, but I feel like they've gotten a lot better at that, like where you can talk to, like, I had to edit a newsletter issue that I put out the other day that had some talk about the... the Trump election that just happened, like Trump winning. And, you know, and I understand that's controversial to people. Some people hate Trump, but I thought that like why it's good for AI. Claude refused to edit the newsletter issue. Right. I was like, that sucks. And then I went to ChatsBT and it not only did it, but I felt like it did as good of a job as Claude would have or better. And so the fact that it's like that you can't, that there's certain things that just won't help you with like long-term, but they keep doing that. It's like, I don't know how I can rely on it as like my main assistant. assistant for my life if they do that. So to add to what you're saying and to actually sort of agree with bumping it down to a tier, I do use it a lot. It is one of my more favorite tools. But here's like another argument for pushing it down to a tier. They jumped, bumped the price on Haiku, right? Because they said, well, it's smarter, so we should cost more, right? So they bumped the price on Haiku. And the other thing about Claude, like the biggest complaint I hear the most is getting rate limited. I don't really hear people talking about how like in chat GPT, they get to a rate limit very often, but in Claude, everybody seems to hit the rate limits. I don't really find I hit the rate limits very often unless I'm writing code. If I'm just like having conversations about like a YouTube script or something like that, I never run into rate limits, but I do know that's like a big complaint people have is that for whatever reason, they run into rate limits a lot with Claude. Yeah. I mean, I think, I mean, they, they, they have to be more cash strapped than opening eye for sure. Like no one talks about that, but. They have to be like, like if you're the number two player in a space, you're always going to have a way harder time raising capital. But they also have the backing of Amazon and Google, right? Like Google is a huge investor in Claude and Amazon basically said, we're going to use Claude as the future Alexa. Right. So it's like, they do have that Amazon backing. Like Amazon could be to anthropic what Microsoft is to chat GPT, you know? Yeah. True. True. So we'll see how that plays out. But I think. I think you've talked to me into sort of conceding it down to an A. All right. So next up is Gemini. The thing about Gemini is we have Gemini and we have Notebook LM on the list. And Notebook LM is powered by Gemini. So if we're just looking at like Gemini on its own, like ignore the fact that Notebook LM is powered by Gemini, we'll rank that one separately. Think about like the Gemini chat and like maybe let's include like Google. AI search, right? Like the AI overviews that Google does. We'll consider that part of Gemini as well. So considering the Gemini chat and considering Google's AI overviews, where would you put Gemini? F tier. You'd put Gemini in F tier? I mean, I feel bad. Like Logan is going to hate me for saying that. And I will concede I have not used it recently. I know everyone said it's gotten a lot better. And so I should use it again. And I just... I feel like I haven't seen a good, like Notebook LM is a great product. So like, that's actually the first great product I've seen from Google in a long time. I feel like the Gemini released, I feel like it's been like a totally botched launch. And I don't know anyone who personally uses it. So that's why I put it F. See, I'm in the boat of like, I hate to put anything in the F tier. Like, because, you know, I actually like a lot of people at Google. The thing about Gemini, there are redeeming factors about Gemini, right? Like Gemini's got the largest context window of any platform, right? So if you want to upload the entire Lord of the Rings book series or the entire Game of Thrones book series and have discussions about it, Gemini is pretty much the only game in town that's going to let you upload that large of a context window. So I think that's sort of a redeeming factor for them as well. Gemini, I believe. also has the ability to sort of read videos now too, right? So you can put a video in there and it understands what's going on in the video and can tell you about what it sees inside of that video. I don't know if any other models can do that. Yeah. Gemini also powers Notebook LM, which, you know, we're kind of trying to keep the two separate at the moment, but like Notebook LM is really, really useful. So I have a hard time like throwing it as like a solid F, but the only reason I would put it sort of lower on the tier list is for the same reasons you mentioned. It's not really one of my go-tos. It's not something where I'm like, okay, I need to go use AI for this. Let's go pull up Gemini. No, I'm usually going to go to Claude Chachi PT or perplexity, right? yeah um so for me it's just not the top of mind tool yet that i would go and use when i need ai's help yeah um but it does have things that yeah it's got great capability like if i was ranking it purely on capabilities i might put it a or b and i'm not a huge fan of the ai overviews yet um like you mentioned it like it kind of didn't know the difference between real news and like memes right like that's why i was telling people to put glue on pizza and eat 10 rocks a day or whatever Right. Like it just it didn't know that like this wasn't factual information. It was just information it found on the Internet. Right. But I think they've improved a lot of that stuff as well. But it's gotten to a point now where I don't really use Google or see AI overviews too often because I go to perplexity for that kind of stuff now. So I don't know. Personally, I have a hard time throwing it as an F because it is a capable model. It's just not one of the models. It's like my go to. I think I'd probably maybe put it like a C or a D. D. Let's do D. Something has to be lower. Otherwise, we're going to have all A, Bs and Cs. Yeah, that's true. That's true. And if I'm looking at the two, I probably still would go to Grok before Gemini as well. Yeah, I would. Yeah. I'm going to jump over to Llama here because this is the other language model. I don't think any of these other ones down here are language. Well, we can do Perplexity. Let's do Perplexity next. So Perplexity for me, solid S tier. I use it every single day. S tier. Yeah, yeah. Like right now, I'm like... Like ChatsBT and Perplexity, that's like daily for me. And they just, you know, I mean, Ervin was one of our first guests who came on. Like, you know, I've talked with him personally. He's a great guy. And they ship so fast, you know. And coming from like Silicon Valley and doing startups, I know like how fast you ship things is so important. I would say they're shipping faster than anyone. Like they're doing better than OpenAI in that regard. And they're more agnostic, right? So like you can like Perplexity powers part of future tools, right? And. on future tools, it uses Lama with Perplexity. So it does a search on the web, finds the information, uses Lama to sort of like write the summaries and stuff. When I use Perplexity like on my desktop, not within the API, but just like I'm going to use it to look up some information, I'm using it with Cloud 3.5 Sonnet. They even added a new feature where you can use the GPT-01 model or the OpenAI-01 model and actually have it. do more like logical thinking through with what it found in its search results. Right. So like it actually allows you to use like whatever model is your favorite plus the web search results, you know? Yeah. Oh yeah. I think that's awesome. That's one of the few that we can totally agree like S tier. That's one of our main go-tos. One of the most useful AI tools available out there right now, in my opinion. That is interesting. So it kind of has replaced not only Google. but Wikipedia as well in a way, which they probably use Wikipedia as a source, but that is interesting. Yeah, yeah, yeah. And I've done a lot of like research. Like the other day, I was looking for a new like headset with a microphone on it, right? Like headphones with a microphone. And I asked it to basically like look for the most recommended headset microphones by YouTubers and rank them like top 10. And it went and looked for all of this and it linked me up to the YouTube videos and it ranked them for me. And it gave me the source. And then I went, okay, cool. This is how it ranked them. Now I'm going to start watching some of these YouTube videos. And I started clicking into the videos to see what the people were actually saying about them, what the sound quality was like and all that kind of stuff. So I was actually starting to like click into all the resources after it did all the research for me. And like, man, it's just, for me, it's been really, really helpful. It's probably the most useful AI tool. Cool. Let's do Lama next. I'm curious your thoughts because I don't think you probably use Lama very often. Yeah, that's what I was about to say. I have a hard time ranking. I'm going to trust your rank a little bit better than mine. I've used it one time. So Lama, like I mentioned, Lama is actually the AI that powers what's going on at Future Tools. So all the descriptions and things like that that are on the website are a combination of Perplexity and then Lama, right? Because Lama is really kind of the cheapest model that you can use right now. Okay. Because it's open source, So like, I think the fact that it's open source, you get, it's got to have some like brownie points for that, right? Like that, that, you know, gives it some cred because so many of the open source models that people are using now, they, they started with Lama and then sort of modified them, forked them, whatever you want to call it and like built their own models. But we're based on Lama originally, right? So I think that's like an important factor. I would probably put it into like B tier. Like I would use Lama probably before I would use Grok or like getting a text response just because it is that open source model and it is insanely inexpensive to use as a large language model. When I use my Meta Ray-Ban glasses, they've got Lama built into them. So if I rarely ask my Ray-Ban glasses questions, but when I do, it's answering with Lama. For me, it's gotten to a point where like most of the large language models are like as good as each other for my use cases. Right. Like a lot of people that use these large language models that need them for like the logic reasons or more like really in-depth scientific complex topics. They're probably a little bit more sensitive to which model they use. For me, if I'm just like asking questions, that's like general knowledge that's out in the world. Or if I'm trying to get it to help me like write a story or if I'm getting it to try to summarize something for me. Most of those use cases, like all of the models do pretty well at this point. And so the fact that we've got an open source version that does what ChatGPT and what Claude does and, you know, what all these other models do really decently for most people's use cases. That's why I think there's value there, right? Like most people are using something like ChatGPT. to help them write their emails better or to cheat on their, you know, social studies homework or, you know, to do fairly simple things that pretty much all the large language models are pretty professional at now. And so, you know, Lama is the open source version of that. Right. Yeah. I mean, since I haven't used it much, I'm not going to like argue on it. I probably would have put it the same as Grok if it was just my own ranking. But since the fact that you're actually using it and like neither one of us daily use Grok on a daily basis. Like, yeah, it should be a B then. Let's move on now to, let's talk about some of these like coding platforms. So we got GitHub Copilot. We got Hercer. Replit, AEI Agent would be another one. Yeah, so we got Replit. There's another one that people are probably gonna be angry that we put on here. I can't remember the name of it. Is it Bolt? Are you talking about Bolt? Oh, is that? Yeah, it is Bolt, isn't it? Bolt is... pretty good. I've only tried it for a couple minutes, so I don't think we should put it on this tier list because like I haven't played with it enough myself. Yeah, but it was impressive. Like I literally told it like I was watching a video earlier today from Tiff and Tech, who's a YouTuber who does tech videos and she actually asked it to make me a clone of Spotify and it did. And I was like like literally that was the prompt. Go make me a clone of Spotify. And like when I was using it, I asked it to make me like an Evernote type tool where I can sort of save bookmarks from around the web in it and it made it pretty quickly like it's not like cursor where it looks like a ide where you're you know you can see the code and all that kind of stuff it looks like a chat bot you ask it to do something and it just sort of runs through the code and it's like all right here's your app i don't do a lot of code so i'm gonna have to lean on you a little bit for these code ones yeah and disclaimer i don't code every day like you know i've done three tech startups i've coded on and off most of my life um but uh You know, GitHub Copilot, originally, I loved it. It's actually how I got involved in AI, or at least more in the current wave of AI. I'd actually been doing some stuff with AI in my previous startup. We'd actually had a computer vision department. But in this current wave, it's how I got interested. It was like trying out GitHub Copilot and seeing how amazing it was. I was actually going back and just for fun, I was like, oh, I should pick up C++ again. and maybe I can make like a cool little game demo or like something I was doing like a Zelda kind of game, but it was kind of like Dark Souls theme, just like messing around. And it was amazing how fast I was able to like learn C++ again just using GitHub Copilot. I was like, oh, because I know basic like coding logic and stuff. And it was like, it was just like helping teach me things so fast. But I haven't used it recently and it seems like less people are using it. It feels like maybe they fell behind a little bit compared to Cursor and other ones. So I think I would probably put it like in a solid, I don't know, B tier would be, you know, where I'd put it. Yeah. Okay. Yeah. I mean, I've never used Copilot, like not even once. So like I really don't have any sort of rebuttal or argument to that. So B tier it is. Yeah. I mean, it's mainly really good for like inline suggestions of like, oh, you're writing, you're writing something. And this is what I think you're wanting to write and give you a great suggestion. And you press tab to complete it. It's like, it's, it's like, that's how they got started. I think they have some other more advanced features, but I haven't heard of anyone actually using them. So, yeah. And then we have Cursor, right? And Cursor, we actually did an episode with Riley Brown where we used Cursor and built a whole app like within the course of that episode. It's pretty damn impressive. Where would you put that one? So, I mean, I would put Cursor at S tier because I think it's actually having a major impact on software development. Like, it's like, there's lots of AI stuff that we talk about where it's like, oh, it's cool and it has promise in the future. I'm hearing from lots of friends in Silicon Valley that, like, lots of teams are using Cursor now. And it's the fact of, yeah, it can give you the inline suggestions like GitHub Copilot does, but also it can, like, look at your entire code base. And... And it can have context of the entire code base. It can, it can, it can, it's helping people a lot for like debugging issues, you know, because like debugging issues is like one of the things when you're in software development, it's like the thing where you're like, damn, why am I doing this job? You feel like the job is amazing when you're like, you're having a concept and you start, you know, you're putting it together and you get to see a demo of it work. But as soon as you start debugging, it's just like, I hate this job. Yeah. Yeah. I'll, I'll keep it up there in S tier. I agree. I, you know. My sort of main experience using Cursor has been watching Riley go back and forth with us on an episode. And within, what, the course of 45 minutes, we went from, like, we didn't have an idea. Like, we used AI to come up with the idea on that episode. And then by the end of the episode, we had a working app. And I was like, this is a 45-hour long episode, and we have software now. Like, that is blowing my mind. Yeah, something else I can do that's amazing that I haven't heard a lot of people talk about is you can... Like a lot of times when you're coding, you might be working with a certain API or some kind of, you know, some kind of library that you're using that you don't fully know how to use it. And you end up having to go back and back to the docs over and over and over. Like, oh, let me look. How do you do this? How do you do that? You can like just like link to docs and you can like like kind of like upload it to a cursor. And like you could say, here's whatever API. And then you can just like at tag that and talking to cursor. And it knows all of the docs. And then you can just ask it any question and get the answer to how you do that. That's like so much better than like opening up another window and like trying to have to search through like, oh, how do you do this? And like learn it all again. They even have stuff too for like the command line where it'll help you like doing command line stuff. It gets really tedious, especially if you don't do it every day. If you do it every day, you learn the commands. It's not bad. But like if you do it like once a month, like me, like you kind of forget some stuff. For sure. It can just like, oh, here's what you're trying to do. And here's the command. Every tool out there has like an API and every API is sort of slightly different from every other API. Yet somehow when you use cursor, it seems to know how to implement the API for the tool you're trying to implement. Right. Like if I'm trying to implement the whisper speech to text feature in an app. Right. It knows how to just use the open API. When I was trying to get it to do like like a WYSIWYG editor. Right. Where I can like, you know, make. fonts bold and stuff in my text, it just automatically knew how to pull in the proper APIs for that kind of stuff. Right. So that to me is also really, really impressive. It's like you don't need to go to whatever tool you're trying to integrate with and learn all about how that API works. It seems like cursor will go and do that work for you to figure out how that API works. Yeah. And I think this is an example, too, of like when people are making an AI startup, they should be looking at cursor and perplexity, like model agnostic. ship fast, solve a specific problem. Yeah. Then we got Replit, which I have used Replit a little bit, but to me, Replit was more of just like that IDE, right? It was more of like a replacement for Visual Studio Code for me, but it was saving it all to the cloud instead of saving it all to my computer. Well, yeah, on that side, it's for deployment, right? They're kind of a competitor with like Vercel and some of these other startups to help you easily deploy websites. But... I'm specifically wanting to rank the Replit AI Agent, which is a newer thing that they have. So I'd probably put it in a C tier right now with the potential to be an S tier. I tried it. It's cool. I mean, at least you go from start to finish. I sat down with my son, and he had a ton of fun. We basically just sat down. We gave it some ideas of what we wanted to build, like a website. He wanted to do something that was Minecraft and League of Legends. And it was like a, it's like a guide for these two or something, you know, and it went from beginning to end and show and shows you the code too. And so you can dive in more and see the actual code it's generating as well. If you want, it takes a long time. Like a lot of times it was taking like three to five minutes to generate the stuff. Um, but it helps you create the entire website and then deploy it. So it's like from beginning to end, the quality was not as good as if I wouldn't, if I would've just used my own template and then probably use cursor to help me and stuff, I probably would've got something better out. But you could see long term where there will be a class of people who, if you have an idea for a startup, probably the best way in like a year from now will be to go to something like the Repl.ai agent. And say, make me this landing page. It's going to make you a beautiful landing page. It'll deploy it for you. The whole thing will be done in 10 minutes. Yeah, yeah, yeah. Right, right. That's going to be probably possible in the next year. Or even just sketch something out on a piece of paper, like scribble out like a rough layout of what you want and then upload that image and say, make me a landing page. Here's a sort of quick template. And then it'll make a beautiful version of what you, you know, you sketched out. Something that we saw, you know. back at the GPT-4 demo, but it's never really panned out to be like what we've seen in the demos yet. Right. Okay, I'll put it there because I can't argue otherwise. Yeah. So let's do like the AI image generators now. Okay. Let's start with probably the most well-known with Midjourney. Where would you rank Midjourney? F tier? yeah yeah totally yeah some of our friends are gonna kill me if i do that um god i have a hard time ranking it like there's like part of me that wants to say s and there's part of me that wants to say a i guess that's kind of like honestly i'll be honest with you i'm gonna struggle with the ai art tools because like i actually use them all for different reasons yeah i mean i used to you know when i first started dabbling with ai art i used stable diffusion because i could run it locally on my machine and i just thought that was so cool um and then i started using mid journey and i was like oh yeah you get you know it's easier to to create beautiful things with it i hated that you had to use it on discord it took them forever to get the web version i thought that was so dumb well i would think mid journey i think mid journey kind of argued that like they were more of like a research company building up a really good art tool and they didn't they were less concerned about the user experience at that point right they were just concerned with making the best image model they could this is our user experience live with it or go somewhere else. We don't really care. We're going to make the best image generator there is. And only recently have they started putting like effort into user experience. Yeah, I'm kind of convinced that the image generation part is going to become a commodity and the user experience is the thing. So I think I would put them at A. I think a lot of people would put them in S, but I think I'd put them at A. I would agree with A, I think. Yeah, I'm not using it as much as I used to. There's great alternatives. I think they ship the website. And the new user experience way too slow, which makes me more pessimistic on them long term. Yeah, I think, you know, if I'm going back to how I was sort of ranking things earlier on in this episode, I was kind of ranking them by like how often I would go to that tool opposed to the other tools. Right. Mid journey isn't really one that I go to is like one of my top, you know, two or three anymore. Maybe it's top three, but it's not top two. Right. Top. two for me would be Leonardo and stable diffusion are the two that I use the most. Um, I don't know. I would probably put mid journey as B tier. Oh really? Okay. Yeah. But I could probably be convinced otherwise. The thing about mid journey is it's like, it's, it's the most well known. It's the one that I feel like at least for a long time, like chat GPT was setting the bar with AI image generation. Like everybody was going, Oh yeah, this one's almost mid journey tier. Oh, have you seen the new version of stable diffusion? it's almost as good as mid journey. Have you seen the new, uh, ideogram? It's almost as good as mid journey, right? Like, so like to me, it gained some points as being that one that sort of been everybody's benchmark of like trying to reach. Um, but I don't really find myself using it as often anymore. And I think it's just because not because I don't think mid journey is very good. It's just because I think some of the other models have gotten. better than mid journey at this point yeah and people people are saying that they're going to move into like ai video and that's what's going on and also they're going to move into like you've created some kind of character persona on mid journey and that can be consistent across different things so i guess you can make like comics and stuff so maybe they will maybe they are ahead in that area and we don't know but also recently sam altman kind of like had this grin on his face when he was talking about how much better ai art in video is going to get soon so i have a feeling that behind the scenes too like all the other stuff that open the eye is building is probably going to give them some edge there too that people are not going to expect just like when we first saw it's like holy crap that's amazing they probably have other things that are amazing that have not been revealed yet if i had to guess the other sort of like negative against mid journey is oh the image generators it's probably the most expensive to use as well yeah I keep canceling it and then resubscribing. I'm like, I am not using it. And I'm like, it's expensive. Then I cancel it. I'm like, oh, I want to use it again. I subscribe. This happens so many times now. Yeah, yeah. So like, you know, for me, because it's on the more expensive end of the AI image generator models, I personally don't find myself going to it as much anymore. That's why I put it there. I feel like. flux might have passed it or at least caught up with it as terms of realism right and when it comes to like cartoony like colorful high contrast like aesthetically pleasing images for me i tend to go to leonardo a little bit more now i will say that i am an advisor for leonardo like i feel like s tier of course leonardo's i need to i need to disclaim that that like i do have equity in leonardo Um, but I legitimately do go and use it more than I use mid journey. Right. Like I do go and I don't think I'm going to put it in S tier. Um, I I'll probably put it in a tier just cause it's the one that I use the most. Um, wait, are we, are we finishing up mid journey though? So let's, so, so I, so I agree with, uh, uh, I had said S or a, but I think you kind of convinced me of B honestly. I think that's like a good spot. Cause like. they were S tier for a while now they're B yeah that makes sense to me yeah and then you know so I was sort of sort of rolling into Leonardo just because I was like referencing it in comparison to Midjourney for me Leonardo is like it still has some of the like issues you might get with stable diffusion where you get some of the like seven finger hands or like um you know like that bird flying up there looks a little weird why does it have three eyes or whatever like you get some of that weird stuff that you get with stable diffusion But aesthetically, like the Leonardo Phoenix model, which is their proprietary model, it's not like one of the stable diffusion models. I think they've gotten it's gotten really, really good. It's like if you look at my YouTube thumbnails, the way I make those thumbnails is I usually generate an image in Leonardo first. And then I take that image and then I pull it into stable diffusion that I have locally installed on my computer. And I do a face swap with local with stable diffusion. So it's like the image was made in Leonardo. And then stable diffusion is what I use to sort of mask out my face and then put my real face on the image that was generated. So I'm using that one constantly. It's what all the thumbnails are made with. But I don't know. Have you have you used Leonardo before? I have like twice. Honestly, it's been a while. So I definitely haven't tried the Phoenix model. My impression was like when I tried it, I was surprised how good it was, to be honest with you, because I felt like I hadn't heard much about it except from. from you honestly like i heard about it a little bit on x but like i was like nowhere near as much as mid journey um and when i tried it when i and this was probably a year ago now so it's probably way better since then um i felt like it was 95 as good mid as mid journey in terms of quality which surprised me i was like oh this is pretty close to mid journey um and i liked the interface a lot better like i thought it was i thought it was i thought it was a cooler app to use honestly um I think I just had the mid journey subscription. So I just ended up not going back to Leonardo. That's honestly like kind of what happened. Um, but it comes to realism. I think mid journey still has Leonardo beat, you know, I think flux and mid journey are better at realism than Leonardo. But when it like, do you remember back when like mid journey was on like version three and they had these like really like vibrant, like contrasting colors and every image you looked had like this mid journey aesthetic that just had like this cool look to it. Right. That's how I feel about the Leonardo Phoenix model is it's like it's got this high contrast HDR sort of like really, you know, the colors really pop. It's got like that aesthetic to it, which to me is really, really pleasing. I feel like with mid journey, like each version that's come out of mid journey, they've sort of moved away from that like mid journey aesthetic. And now they just kind of look like, you know, whatever you're prompting, like whatever style you're prompting. Leonardo to me still has like a style, I guess. Yeah. So you are using it for your thumbnails on a daily basis. Is that right? Well, I don't put out a new video every single day. Not daily, but every week. Every video I put out, yeah, I use Leonardo for the thumbnail. Yeah. Yeah. I'm kind of like, I feel like I could see A, but also like people are going to freak out about that saying Leonardo over mid journey. And also since you're an advisor, I'm like, damn, I should probably put it as a B. I'll put it as a B just as a precautionary measure. I, you know, so here, just kind of thinking it through here, right? I think mid journey and Leonardo should be kind of on the same level. Maybe like if we're ranking them like this way too. So like, this is a higher B maybe, but like, they're probably about as good as each other. Like I would go to mid journey. If I need something a little bit more realistic, I would go to Leonardo. If I'm looking for that sort of like aesthetic, the color's going to pop. It's going to make a good thumbnail kind of thing. Yeah. But they're probably on par with each other. Okay. I can see that. All right, so I put stability AI here, but this is meant to be for stable diffusion. Stable diffusion itself doesn't really have a logo and everybody kind of like associates stability AI with stable diffusion. I mean, so for me, stable diffusion, I'm probably just going to put it right alongside these other ones. um because i use it just as much as i use leonardo right like that's what i use for like my face swapping maybe i put it as a c tier actually just because like i don't use it to generate images like almost ever i use it to do like in painting so i generate the image with a mid journey or a leonardo or like one of those tools first and then the only thing that i'm using stable diffusion for at this point is to go in mask out my face and then like superimpose my face using ai so i can't I don't know. I don't think I can put it on the same level as these other ones. But I was going to say F, which is like probably get me killed by like everybody who. But I mean, I feel like they dropped the ball so much. Like early on. You're referring to Stability AI, the company. Yeah. OK. Yeah. The stable diffusion is like an open source image generation model. Stability AI has made a handful of the weights that have been made available, like SDXL, SD. 2.0, SD 3.0, like those were made by Stability AI, but Stable Diffusion itself was an open source model that was built before Stability AI even existed. Right, which was super like, I feel like there wasn't a whole lot of transparency around that. So yeah, I'm probably judging the company, not the model itself, because like I have a lot of feelings about the company in terms of the, you know, statements that were made to investors and other things. When SD, was it? 3.0 came out and everybody was generating the images of people like laying in the grass and it would have like these mutated people that had like only legs and no head and you know random stuff like that yeah yeah yeah and i guess i was also like you know early on i was using stable diffusion a lot uh and i just i personally don't use it anymore but it sounds like you do i actually have a friend who has a company as well and i know he like i did he has like a discord bot or something like that i know he uses it so it sounds like people are using it so probably yeah it's probably not an f Yeah, well, I guess if we're looking at that as a whole, right, like most of the models that people are going to use for stable diffusion today are the models that were generated by Stability AI. So I guess like you can kind of lump it in, you know. Well, just as we've been talking about this, I've already knocked it down two tiers. Okay, okay. Yeah, I think D seems good. Like it's going to be controversial. Some people are going to say it should be like an A. I don't think anyone's going to say S. Well, again, my logic on it is like how often do I find myself going to it when I need to generate stuff? And I do use it often, but I don't use it to generate the images. Like the only value I'm getting from stable diffusion right now is that face swapping feature. Okay. D, let's leave it at D. We have to have something that's lower. I mean, like. Yeah, but looking at what's left, what would go lower? Yeah, yeah, that's true. That's true. We may not get anything at F. Like, we'll just like. We're just too nice. Yeah, we don't want to burn any bridges with any companies completely. So it's like hard to throw anything below a D. stability AI of the company F. All right. But stable diffusion itself, let's leave it in D because it is open source. You can literally generate anything if you're running into like roadblocks with censorship or whatever with any of the other AI models. All right. So this next one here, this is playground AI. Quite honestly, I haven't used it a ton, but I did use it recently. They've sort of rebranded a little bit to being like. the AI generation tool for like graphic designers, right? Like they've, they've made it really good at creating almost like vector style art. Like it's not actually generating vectors, but it's generating like vector style art, like logo type stuff and stuff that would be good on like t-shirts and mugs and stuff like that. It seems like they've kind of gone more that direction and sort of niched into it, right? Like you've got like scenario dot GG, right? They've really, really like leaned into like game assets where playground, they've really seemed to lean into like where, what you go to, if you want to create logos or graphics for like your, you know, your company assets and icons and things like that. And for that use case, it seems pretty dang good, but I haven't used it a ton. Like that was me playing around with it for like 30 minutes or so. Having a hard time ranking this one for like multiple reasons. Like I, I haven't tried it recently. I tried it before. Haven't tried it recently. Next time we do this, I probably should try everything before we do this. And, you know, the founder, Suhail, I have a lot of mutual friends with him in Silicon Valley. He created Mixpanel back in the day, which was like, was for a while considered like the coolest analytics tool. Like instead of using Google Analytics, you use Mixpanel if you wanted more details and stuff that mattered to you. So, you know, he's definitely a great founder. But when I did try it, like the user experience and interface was like really good. It was like. That and Leonardo were like two of the ones I thought were pretty cool. I mean, just kind of going back on my sort of fallback, like. logic with how I've ranked everything else. I don't really go to that one very often. I mean, I use stable diffusion more than I use playground. Yeah. But I also think that this should probably still rank above stable diffusion because as far as like people listening to this episode, it's a lot more user friendly than stable diffusion. Like you're probably more likely to go like, unless you're using it for a very niche use case, like I am with stable diffusion where I'm going and face swapping for YouTube thumbnails. You're probably not going to use stable diffusion, right? You're probably going to. prefer something like playground, which has stable diffusion models built into it. So you can just select the model and get playground to generate the image with stable diffusion for you. Right. Also it's similar to like the AI version, uh, or the art AI art version of like a perplexity or a cursor where you can go into playground. You can use playground model V3, but you can also choose Dolly three, or you can also choose stable diffusion, uh, S SDXL or SD 3.5. Like that. You could go and pick whatever model you want to use, or you can use their own proprietary built-in model. So I think that gives it more value than Stable Diffusion right off the bat because it does what Stable Diffusion does plus, you know? Yeah, and anyone could listen to this and just go instantly try it and get the gist of it in a few minutes. But as far as the images that it generates, I don't think I'd put them on the same level as Leonardo and Midjourney. I don't think it's up to par with those. quite like it's close but it's not quite as good as those like with the quality of image yeah the last time i saw images from it it seemed like kind of closer to like stable diffusion level yeah yeah and then we've got um just moving on we've got adobe firefly this might be my first e-tier here's my problem with adobe firefly out of all the image generation models it is the absolute hands down most censored right like i tried to get it to generate an image of the eiffel tower And it said, no, we can't generate that for you. And I went and looked it up. And apparently images of the Eiffel Tower are technically trademarked or copyrighted or something like that. Like it wouldn't generate an image of the Eiffel Tower. When I get it to generate images of actual people, it's not great. Like the realism isn't quite there when it comes to like aesthetics and generating images that like I just think are aesthetically pleasing with like the color palette and the contrast and stuff like that. Never very impressed. Like every time I've gone to Adobe Firefly to go try it to see if like, okay, has this one gotten better than it used to be? It doesn't really like the UI has gotten a lot better. Like the UI, I have no beef with the UI. I think it's a decent UI. I just really, really don't like the images it generates. And it is so insanely censored that like I'm always sitting here banging on my desk. I was so frustrated that some of the images it wouldn't generate. Yeah, so I only saw examples of it then. I haven't tried. I've used generative fill and some of the other Photoshop features they rolled out, which are pretty cool. But yeah, it sounds pretty bad. I didn't see any examples that looked great either, like on Twitter or anything. There was nothing that was like, oh, that's awesome. And to think about Photoshop, it's a creative tool. It's like, how can you put such a restriction on a tool that creators use on a daily basis? It just... It's... there's no logic there for that because like you know maybe they're maybe they're doing it for legal reasons but like it's like saying that you've got a hammer and you can't use a hammer a certain way it's like well that's why would i buy that hammer it's like you know well you know it's it's uh like you can use photoshop to draw anything right like so well you're bringing up some some points here that actually want to touch on like adobe claims to be like the only ai like ethical image model right because they've only trained on like images that are inside of adobe stock and like public domain photos, right? So they claim that like, you're not generating on a model that was trained on like everybody else's artwork because they had the license to use all of the images in their training data. So the people that are really, really concerned about the like the ethics of how the models were trained would probably go to Adobe over the others because of that sort of like ethical, you know, piece that they put on there. Also, you mentioned that Firefly in Photoshop, that has actually been useful. Like I kind of forgot about the fact that it's actually using Firefly when you're using Photoshop. And I do every once in a while use Firefly to like fix stuff, right? Like I'll generate an image with Leonardo or Midjourney or something like that. And there'll be like a weird artifact in it. And I'll pull it into Photoshop, circle that little weird artifact, click the generate generative fill button, and it'll like remove that for me. And it works at that really, really well. But even saying that, the censorship is so bizarre sometimes. I honestly think this should be our first F tier because... like a creative tool that doesn't allow you to create things you want to create. And also, I haven't seen anyone sharing good examples of using it. I haven't heard of anyone actually using it. Everybody uses it in generative fill, though. Like if you're using Photoshop with generative fill, that's where you're most likely going to use it, right? Yeah. And that's where I find myself using it from time to time. But even there, like I've had times where like it would be my face in an image and I would circle my eyes and say, like, put sunglasses on me. And it would say like, oh, our guidelines don't allow that. Like sometimes the guidelines just... present false positives or false negatives or you know whatever direction that should go it it thinks you're trying to generate something that you're not actually trying to generate right and sometimes it's just like really weird like that or like like i've had issues where like i had like a six fingers on one of the hands that it generated right and i was like circle one of the fingers and it'll be like oh sorry this is against our guidelines we can't we can't change this image because it obviously thinks the finger is like you know a different appendage or something i and so like you run into that kind of issue with it But I have a hard time putting in F tier because I do find it useful. But we should treat it as like it's standalone versus the generative feel. Like if we had generative feel, maybe we're talking like a B tier or something like that feature in Photoshop. But for like a standalone image generator, I think we have to judge it that way to get like an accurate judgment on it. And I do think that would be an F. Yeah. You know, and another negative for Adobe, too, is like they're. pricing model has always felt a little predatory right like where they do the whole thing where it's like it's this much per month but you have to agree to a year-long contract to get it at this much per month and then people don't use the tool anymore and they can't get out of there it's just that to me feels very icky and i've always hated that about adobe i yeah i've there's been once or twice where like unintentionally i end up paying them like 500 bucks or something yeah it's like that sucks all right so next up is magnific I haven't actually used Magnific to generate images, but I've used their upscaler quite a bit. And I think it's pretty cool. But I think the big complaint about Magnific is the cost, right? A lot of people say it's like for an upscaler, it's like kind of pricey for that. But what are your thoughts on it? Yeah, I got free credits. That's my problem as well, is they did give me some credits so I could play around with it. But whenever I've talked about it on a YouTube video or something like that, people are always like Magnific. prices is crazy you know right i i haven't used it recently so i'm gonna have a hard time giving it like a really high ranking uh but when i used it it seemed great like like upscaling was always one of the issues like you'd use different ai art tools and then like there was it was crazy that there was no good upscalers just like why is it how is this still an issue and then the fact that and then it was kind of cool how it would actually make them look better like some stuff you put in there it all of a sudden look look nicer which sometimes that was bad too, because sometimes it would actually change the essence of the image as well. So I had a hard time kind of like playing with the tweaks on that and stuff. Like, how do you get that right? It upscales, but it also hallucinates. Like it hallucinates on purpose, right? Like it will add extra stuff that wasn't in the original image, but it creates this like cool effect. Like I've always thought it looked really cool. Like we've seen those things that circle on Twitter and stuff where it's like Laura Croft from like the original Tomb Raider game and somebody upscaled it with Magnific and then it looks like a... like you know a triple a game graphics i wouldn't say it looks like ultra realistic but it looks like modern day game graphics with the upscale done to it like some of that kind of stuff is really really cool but to me it falls under the category of like novelty like i don't yeah use it a ton yeah i thought you were gonna say like the uh the new lara croft and then change it to more like the original lara croft but i won't you can go that direction if you want as well but um uh yeah i don't know i'm feeling like like probably um Jeez, where would we put it? Should we put it like C or B? See, I'm going B or C. I'm probably leaning more towards C because if I'm looking at like the value of the tools in B, I don't think I'd put Magnificent like the same level of value of these tools. And I have to like disclose, we probably both have talked to, what's his name? Javi? Is that how you say it? Javi, yeah, yeah. Javi, yeah. I've talked to him several times on Twitter. Nice guy. I was kind of shocked that he built this because like I just knew him as like this guy on Twitter. And it was like awesome that he built this amazing startup. So it's like props to him, but he's already sold the company now. So now it's owned by FreePic, which FreePic is an AI image generator that also has Magnific built into it. Yeah, so we can talk more crap about it since he's already sold it. No, so I would say C because it's expensive and just long-term, I don't see the value long-term just because it seems like, you know, and I think I may have even said this to him. yeah you should sell it like if you can because like it took off like crazy fast and it's like but like long term how does that become its own company or own product like it's an upscaler like a cool upscaler like yeah that should be part of mid-journey or leonardo there is one tool that we're missing um Because if we're talking about AI art, you've got to have Luma in here as well. I mean, sorry, AI video. These are AI video tools right here. So we've kind of covered the AI art. Now we're sort of shifting into AI video here. We've got Runway, Pika, Kling, and Luma. Because Luma's got Dream Machine. And Dream Machine was pretty damn groundbreaking when they showed it off. But let's start with Runway. For me, Runway... For me, runway is nest here. And my reasoning for that is like runway just ships, man. Like in the last couple of weeks, we saw runway ship act one, which was like the sort of lip sync. You can and it follows the emotions of a character. So I can make a video where I'm talking and I'm moving my eyes around. And every time I blink, the animation on the video blinks. It does all that. And then they released the multiple camera angles where you can make a video from like. any angle of an image that you upload in there and they um and that's really cool and then they've got like frame interpolation for like images where you can take one image and another image and have it like frame interpolate between them and do this like morphing effect and they've got the ability to go and like remove the background of any video and turn it into like a green screen video and like to me when it comes to ai video like every week runway's got some sort of cool new feature they just they just ship and they ship and they ship And the generated videos that come out of Runway are really impressive to me. When I'm going to generate an AI generated video, I'm either using Runway or Luma most of the time. And for me, I'd put it in S tier. Yeah, I get the S tier. I just don't think it's on the same level as those other three companies in terms of like actual value that can be created with it. Like as of right now. That makes sense. And I'm biased because I do a lot with video, right? So like I'm playing with video tools a lot. I'm sure a lot of people that listen or watch this, you know, this podcast don't do AI video as much as I do. But when it comes to the AI video platforms that are out there, to me, Runway always seems to be like one step ahead. Again, we don't know where open AI is with video like they haven't released it to the public. You know, we talked to Don Allen Stevenson, the third on our podcast, who is somebody who actually got to use Sora and. basically he was like i can't give you more details but it's better than you think right so there's good stuff coming out of open ai in that video front at some point but yeah as far as tools we actually have access to and can play with right now i think runway sort of leading the pack with ai video at the moment but i think a makes sense all right so next up you got pika there's no way i'm putting pika in the same like realm as runway um Pika generates AI generated videos as well. The videos are not great. Like I don't really feel like they've got great realism, but their new sort of thing is like, you know, you see the videos where like, there's a giant like compressor that compresses somebody or like the, is it cake kind of thing where a knife comes in and like slices, whatever you're looking at, it makes it look like it's cake or like they got that like balloon feature where it'll blow up really big and then like float away. And it's got like all those kind of like, gimmicky like uh like yeah effects that you can put on videos and that's kind of fun to play with there's not a whole lot of use out of it out of like outside of like memes but right kind of fun yeah yeah and when they first came out i got i got like a demo i'm not sure if you did but like a few like ai influencers got a demo before other people and it was at that at that moment like at like for whatever reason like cartoon kind of characters it seemed pretty good it seemed like better than some of the other models for like that kind of art that kind of style but then it seems like there's been so many other things there's been all these models from china cling and others that are like probably better looking for that kind of stuff the special effects look cool maybe that's what they'll keep doing maybe it'll be like the special effects viral memes kind of company but it seems like that's expensive like are people going to really pay for that like you know um that's where i run into like a little bit of struggle with like ranking some of this stuff is it's like i think it's really cool and it's fun to play with but i don't know if i see the real world use case where like if you're paying for it you're gonna get roi on it you know what i mean yeah it does seem like that kind of puts it in like the c though yeah i think i agree with c i think i think it's kind of like a novelty and you know like realistically most tools should fall under a c right like that's that should be like our average if it's like a bell curve right so yeah it's cool it's it's cool it's uh it's fun to use uh it's not the best one out there but it's it's not bad i am skeptical about the company long term just because i don't know how they build a business around that because like with runway i can see it like if you if you get baked into hollywood and stuff obviously like like i like like i've told you i dealt with hollywood a little bit like they spend so much money on special effects like some of these films it's like 100 million dollars on special effects it's like well if you could cut that down to a million dollars or whatever you know that's that's huge um and runway was founded by like legit like ai pioneers like one of the founders i don't remember exactly like names here but like one of the founders of runway was one of the original developers of stable diffusion yeah yeah so i i think c makes sense and then you've got cling which I haven't really used that much. It's one of the few AI video tools that I personally haven't played with a ton. I've played with it a little bit. I have noticed, like, you know, I used to do those AI video threads, and I have noticed that it seems like right now, though, most of the videos people are putting out on X, a lot of them are Kling, actually. Yeah. So it seems like in terms of visual quality, they are in the realm of runway, is what it seems like. It seems like they're in the realm of runway, but with, like, like the editing tools not being as good as Runway. So maybe like a B tier? It seems like a B, because it seems like they're putting out better videos than Pika. Well, it also seems to me like Ling is a lot more uncensored, right? Like if you went to Runway and said, generate a video of Will Smith eating spaghetti, right? Like the very classic meme that's been going around for years now. It'll say we can't do that because we're not going to make videos of Will Smith. I believe Kling will. I think Kling is like... We'll make whatever you want to generate. Things from China, they'll do whatever you want. Yeah. And that is one thing I've seen people kind of talk about online recently. Like, why is... And there's one or two other Chinese models, too. We probably should be listing them, but they're also pretty good in AI video. And it's like, why are they so good? Is it like Halu AI or something? I don't know how to pronounce that one. Exactly. Yeah, that's why it's so hard to say that. I couldn't really say it. But those are all pretty good. Those are all probably better than Pika right now. Or in the ballpark of it. Pika's got the cool special effects, but they're in the ballpark of being somewhere in between Runway and Pika, which is pretty impressive. They may end up having some of the best models out there long-term. Yeah, I agree. I think B is the right place for them. I think they live a little bit above Pika, but not quite Runway level. Yeah. So then you've got Luma, and Luma has their Dream Machine, which is probably about on par with what Runway... does as far as like the video generation, not nearly as many features as runway has, but Luma's the cool thing about Luma dream machine that I really like is that you can give it two images. You can give it a starting frame and an ending frame, and then it figures out how to make a video that goes between those two frames. So for example, if I put an image of myself as the starting image and an image of like a wolf howling at the moon. as like the other image and i want it to like animate as me morphing into a wolf howling at the moon it can do that and like that's where i feel like luma really stands out is the start frame end frame create an animation between the two frames feature i really really think that feature is cool um it doesn't put it on par with runway but i think it puts it a step ahead of pika maybe yeah probably a b then they get yeah i play with it one time i liked it better Well, the other thing about Luma is Luma was really one of the companies that led the way with nerfs and Gaussian splatting and stuff like that, right? Like, if you have the Luma app on your phone, you could go, like, take an object in your room and, like, scan it in with your phone and create a 3D version with the Luma app. So not only does it have a video generator, it also can scan in real-world objects and convert them into nerfs and Gaussian splats and stuff like that, which... I know is getting really technical in the weeds, but if you're really into like creating 3D assets and stuff, Luma is also capable of that. Right. Yeah, I think B makes sense because it seems like in a way they're going to be competing probably directly with Pika then too, like on special effects and things like that. And it seems like they're probably already ahead of them. Yeah, yeah. All right. So next we got Descript. And I know you haven't really played with this one, so I'll just talk about this real quick. But Descript is a tool where you can upload any audio or any video file. It will automatically... use AI to transcribe that whole file. And then you can edit the audio or video file by editing the text, right? So if there's like a sentence that I want to remove from an audio or a sentence that I want to remove from a video, I just delete that sentence in the transcript and it automatically edits the audio file or the video file to, you know, make that edit for me. And I found it really handy for, um, I use it as like my main transcription tool. So like whenever I release a short form video on YouTube shorts or Instagram reels or a place like that, I always record the video, pull it into Descript, get the transcription and then use that transcription for the subtitles on the video. There might be better ways to do it. In fact, I'm sure there are better ways to do it. But for me, it's just like it's become a really, really simple way to transcribe every video that I make and then make captions on that video. Um, and it, it uses AI to do that all. Is it like something that's life-changing? I don't know. I think it's really, really helpful if you make a lot of video or audio content. Um, but I'd probably put it like middle of the pack. You, you, you, you, can you edit the video with it as well or no? Okay. But you just don't use that feature. I don't use that feature. No, no. I mean, cause I have editors that I hire. I like, yeah. And I use it. I use a tool called time bolt, which isn't technically an AI tool to do like a lot of the, um, like cutting out the gaps in the audio, but Descript will do that as well, uh, to cut out us and ums. I also use Timebolt for that, but Descript will do that as well. I use Descript a lot for just like getting transcriptions. Another thing Descript was like, even before 11 labs came out, Descript was the first tool that I ever saw that can clone your voice. So like if you had a podcast and you misspoke during the podcast, and you wanted to go and change like one sentence, it actually, you can trade it on your voice, go and type in the new sentence you wanted it to say, and it would replace the original sentence in the audio version of that with you saying it. Now, 11 Labs does it so much better. The Descript version is still fairly mechanical AI sounding when you do it, but it was the first tool I ever remember seeing that had like voice cloning built into it. Yeah. I guess I can't argue with that. I guess, see, like the editing sounds way more interesting to me because like the transcript thing seems like that'll just be like, so like you'll be able to do that with like anything in the future. Like I can imagine even YouTube will probably just have all that. Well, YouTube does have transcription built in too, but it's like, you got to upload the video first and I need the transcription before the video is uploaded, you know? Right. But yeah, I mean, I think like transcription is going to be very commoditized, you know, like you've got the Whisper API from OpenAI. It's really cheap to use. So it's really easy for tools build in transcription. But it's somewhat useful for you today. So like seed makes it's somewhat useful for me today. But I also think like if I was if I didn't already have like workflows in place and systems in place for creating and editing my videos, I would probably use Descript a lot more for editing the videos. It's just like I haven't really wanted to change the workflows that I have. So I think anybody that's trying to get into podcasting and they're going to edit their own stuff or anybody that's trying to get into like creating video content and they want to edit their own stuff. I really think Descript is worthwhile for like speeding up the editing process as well. All right. We got three more. Now we're getting into like the audio apps. Yeah. So we got Suno first. I love Suno. Like Suno is so much fun to me. Yeah. I can't put it in last year though. It's too much of a novelty. Yeah. Yeah. I love it. I've only used it twice. I think you've used it more than me, right? I've made hundreds of songs with it. Oh, that's awesome. And you play music or used to, right? So that makes sense. Yeah, I'm a musician. I'll use songs that I make in Suno instead of YouTube videos. Like if you want background music and you don't want to worry about copyright, you can make Suno make you a song without any lyrics and use that as your background music. So you can say I need some like calm, emotional music for this part of the video. And then I need upbeat, exciting music for this part of the video. And you can tell it to not generate lyrics and it will generate like a sound bed for you for your videos. I've also made like montages where I have like a whole bunch of clips from an event that I went to. And then I actually typed in the lyrics that I wanted. Like, oh, now I'm looking at, you know, now I'm looking at the Insta360 cameras on the wall over here. Those are pretty cool. Look over here. It's a really cool futuristic car. And look at that. It's a flying car. And like I would put all these like random things that I was like, basically what I'm seeing in the video. I would put that as the song lyrics, make a song and then make. the montage video like sync up to the song where the lyrics of the song are actually what i was looking at in the video and you can do like really fun stuff like that that makes for really engaging videos so it's fun and i use it and most of the time i'm using it i'm using it just to screw around and just like play with it it's so much fun i used it to uh make a song uh to my wife when she was my fiance and it was just like a love song to her and like i even like i wrote down like our like story of like meeting in kyoto and stuff and it made this awesome song you know yeah yeah uh so it seems like it's like it's amazing technology it's uh it is a novelty it's kind of like it's just kind of fun to use but you do get some real use out of it so it seems like that's probably like what a or b something like that yeah i'd i'd probably have to put it in a b just like going back to like how often i use it for like actionable, practical use cases. Yeah. And not very often. I'm using it for fun and screwing around more often than anything. Would I put it on the same par with Claude? That would be tough, right? If we're saying Claude's an A, I don't think I can put Suno at the same year as Claude. All right. And then you got 11 labs. So anybody who doesn't know what 11 labs is, it's a voice cloning tool. You can train it on your own voice. You can train it on anybody's voice. You type in words and it will speak out in a very realistic tone. It's pretty valuable. And they keep on adding like features, right? Like they've got the 11 labs reader now where you can load in blog posts or PDFs and it will like read it in your voice or it'll read it in whatever voice you choose. And so you could go out for a walk and turn like a PDF for a news article or something like that into almost like a podcast that you listen to. It sounds like a real human and not like an AI robot talking to you. So that's, you know, that's useful. It also creates sound effects now too. So you can say like, I need a clang sound effect, or I need like a, a dog barking sound effect or a car crash sound effect or whatever. Right. And it will create that sound effect for you. So it'll create talking audio. It'll create sound effect audio. It can be used to like make audio from, from text files. It like, there's a lot that you can do with it. Were they the ones too who created the whole persona thing or was that that's who know I feel like there was one word that You can like create like a personality now and like you kind of like save that personality. That's you know just created Yeah, I was getting confused okay That's great with 11 labs You can train your own voice into it and then they even have a marketplace where you can sell access to your voice if you want In fact Matt vid pro he actually put his voice in the marketplace and then started seeing ads online with his own voice promoting a product. And he's like, I should take my voice off the marketplace. Right. It's probably worth more money than that too, like long-term. But yeah, especially if it's like promoting the wrong thing, right? It's like, here are these pills that help you, you know. I'm Matt Fitzgerald. Here's why you should buy Blue Chew. Yeah. Rhino pills or whatever. Yeah. Here's my gut feeling, and we can debate it a little bit. For me, 11 Lives is A tier. I was going to say A is what I was going to say, because like it's actually being used by people. Like there's people in certain industries that are actually using it right now. Yes, like a lot, a lot. Like people are using it for doing voiceovers on like sales videos. People are using it for the faceless channels. Like it's really, really popular among like the faceless YouTubers, right? That just put out content. You can create voices that never existed before. You could say, I want it to sound like a, you know, an 80 year old grandpa who, um, you know, stepped on attack or whatever. Like, yeah, like random prompt. Do we want to reorder these at all? Like, do we want to say like, oh, it should be ahead of runway because people are actually using it? Or does that not matter? Do we know? Yeah, maybe it does go ahead of runway. And also when it comes to like voice AI. So I actually interviewed one of the founders of Eleven Labs for like an NVIDIA panel that we did a while ago. And when it comes to like the voice, the speech AI. like 11 labs is leading the way, right? Like there's a lot of other companies out there that are like voice clone, like AI voice tools, but like 11 labs is the chat GPT of voice tools, right? It's the one that everybody's trying to catch up with. It's the one that everybody says we're almost as good as 11 labs, right? Like they're, they're like, they're, they're, they're the one everybody's, they're the, they set the bar, right? Right. I mean, the fact that so many people are using it in Facebook channels, that's like awesome. That's like an AI tool that's actually being used that changed an industry, right? So definitely makes sense. And then the last one, I'm not sure if this is the proper icon or not, but it'll have to do the trick for this one, is Notebook LM, which as we talked about at the beginning of this uses Gemini underneath. But we're more specifically talking about like the feature where you can upload a whole bunch of documents or YouTube transcripts or links to articles. And then you can. ask the questions and like actually have a chat with all those documents and it will also create like an audio narrated podcast where it actually sounds like two hosts talking back and forth about the topic which i think is really cool as well i have mixed feelings so i know this is gonna be tough i think i i think it's an awesome technology it's it's a great product um you I can see so much potential for it, especially in learning and things like that. You want to learn a subject, you can make a little podcast and you can listen to it. That's awesome. It's entertaining. My son was like, I don't think I've ever seen him laugh as much as when we created a podcast talking about, oh, by the way, the Earth has just been invaded and one of you is actually probably an alien. My son just thought this was so hilarious, listening to them kind of talk that out. Like, by the way, maybe it's you, you know, kind of thing. Or like, this is weird. I'm just really skeptical about Google's ability to make it mainstream or to release new products in general and have them become popular. I'm skeptical of that because I haven't seen it. So I don't know. It's probably A or B, though, I guess. Something like that. Yeah. It's one of the most amazing things I've seen in the last six months. Definitely, I think it's the most impressive thing that has come out of Google in a while. Gemini has always felt like it was playing catch up with chat GPT. It has the bigger context window, but other than that, it kind of felt like it's been playing catch up. But then when Notebook LM came out, it was like using Gemini underneath and everybody went, this is a cool use case. This is a really, really cool way to like. consume information and i'm i'm using it quite a bit right like i'll pull in a whole bunch of like news articles to try to understand what's going on with something and have them explain it to me saying that i do know it has some issues like it will get confused about things sometimes where if you give it like two different news articles it will say like i'm having a hard time thinking of an example but it will like say something happened in one of the news articles that didn't actually happen because it got confused with the wording between the two news articles so sometimes i know it can get confused and the speakers on the podcast actually give the wrong information because of its confusion between all the documents you uploaded um yeah i i do i do wonder there was some recent uh research i think it came from what's his name that guy ethan the professor at wharton i think he's on x I think he shared it showing that there's new data out showing that the more restricted a model is, like the more restrictions they put on it in terms of being more censored and things like that, the less intelligent the model is. I believe it. Because you add all these different layers of complexity there that just makes it less intelligent. So I do wonder if that's like one of the problems they're having, like Google's having behind the scenes in general, just because of the culture of Google or whatever, is that it's all more restricted. So it just ends up being less intelligent, even though it's got great... technology and a huge context window and all that. And I do wonder, it feels like the last great product that Google released that I can think of is Gmail. Right? And Google Analytics, which they killed, basically. Yeah, yeah. Are you talking about products that Google built? Because, I mean, they've acquired some good companies. Oh, YouTube. Yeah, of course. They acquired. Yeah. I mean, YouTube, you've got Google, you know, the whole Google Drive, Google Docs, Google Sheets, all that like whole suite of tools. I use those constantly. Man, Jim and I just feels like Google Plus to me. And so I can just feel like, you know, where they just copied another company. And Notebook LM is not that. It's like it's unique. So I want to like get a decent ranking. I just feel like they it's not there yet. Like it's like has so much potential and they've shown the potential. And I'm not sure they're I'm not sure they're going to be the one to actually capture the value from that potential. Yeah. I also ran into some issues when I was trying to create some content, like they just added a new feature that I was really excited about where you can give some additional context and sort of steer the conversation. Right. Like there's a new like customize button. And I press the customize button and I said, hey, you're a podcast called. I don't remember what I called it, but, you know, you're a podcast called The Next Wave. When when your episode starts, say, hey, welcome to The Next Wave and then move into the rest of the conversation. Right. That wasn't what I was calling it. That's just for example purposes. And it ignored my instructions completely, right? I also gave it instructions one time where I gave it like transcripts from like three different videos that were interviews with Casey Neistat, right? And I said, in the custom instructions, pull out what Casey Neistat does different than other YouTubers and make a podcast about what makes him different and so much more successful than so many other YouTubers. Totally ignored my instructions and just like, talked about Casey without specifically honing in on the things I asked it to hone in on. I almost felt like the customize button that they put into notebook LM is pointless. Like right now, like for me, I couldn't get it to work. It like, it didn't follow my extra instructions. So that was interesting to me as well. Um, yeah, it does make me wonder that that's like, I'm like a possibly a bad product decision. So that, and again, that's like my criticism of Google in general. So like, maybe this is like some great research that turned. turned into like a really simple product, but can they take it to anything beyond that? Yeah. Well, like it was originally Project Tailwind, right? And Project Tailwind was designed to like allow you to chat with all of your documents that you put inside of like a specific folder inside of Google Drive or whatever. Right. And that's what this came in, became, that's what this became. The podcast thing was, I think, just like an extra fun, little like novelty feature that they're like, oh, this is kind of cool. Let's put that on top of it. And it turned out to be the feature that people like the most from it. And so I think that's kind of how that played out. Yeah, so in terms of like, it's like one of the most impressive things I've seen in the last six months. There was like part of me that wanted to put it like as A, but I do feel like maybe it's B or C just because of, I just don't know what they're going to do with it. Like, I don't know if it'll actually become a big thing or not. It feels like maybe it's already like it's got that novelty. It's been cool. Like, especially some of the videos were like, they were surprised where people told it like, oh, you are AI, by the way. And they freaked out about it. Like that was an amazing viral clip. But what they do with it beyond that, I'm not sure. Like, I'm not sure if they'll actually improve it, make it a big. Yeah, I'm leaning towards B, I think. I have a hard time putting an A tier. Like it's not on the same level as Claude, 11 Labs and Runway for me. Yeah, C theme's too mean because it is a pretty cool product. Yeah, I'm definitely using it more than Grok or Replit or any of the tools that we've got down there. So I don't know. B feels right to me. Yeah, B with a potential for S. So if they actually make it better and do a good job with the product, now it's kind of hidden on the site. It's so hard to even find. They need to launch it as a standalone product and really nail some kind of use case. maybe education or something like this like you got a knowledge base at your company now make some stuff that explains stuff about your company to people something they need to nail some kind of use case and launch a product here's what i can see and i'm just sort of shooting from the hip here but like a combination of what perplexity does and what notebook lm does but like in like a conversational format right like let's say i want to go for a walk with my dog or something What would be really cool is to open up a tool that's like this notebook LM perplexity hybrid and just say, hey, today I want to learn about quantum physics. Give me a podcast about quantum physics. It goes and does the like sort of perplexity thing of hunting down all the articles and information you need and then pulls them into the notebook LM side of things and turns it into that podcast. So now all I'm doing is giving it an audio command of like, here's what I want to learn about. And then it gives me a whole podcast after doing the research for me. Yeah, that seems like it should exist. But that's why perplexity is S tier because they're more likely to do that than Google is. Yeah, I agree. And you know, all right. So this is like, that's all the tools we came up with. I know there's tools that we're missing. I know like there's probably a lot of people that disagree. Like I'm actually looking forward to seeing the comments on this episode because I want to see people go, I can't believe we put that in D tier or whatever. I think that's going to be really fun to see and debate. But at the end of the day, I think every single tool that we put on this list has the potential to jump to S tier at some point, also has the potential to fall to F tier at some point. I think, you know, like, this isn't anything personal about the companies. Like, I actually, for the most part, like all of these companies and have messed with the tools and think what they're doing is really, really cool. So there's like, you know, I'm trying to save face here a little bit and be like, hey, don't hate us for putting you in this tier. But, you know, I do think all of these tools have the ability to like move up and down. This is sort of more of a ranking of like, how useful they are to us in our lives, which is obviously very subjective right now. So, you know, just want to kind of disclaim that before we wrap this one up. But I don't know. This was kind of fun. I think this is going to be our longest episode we've ever put out, but it was really fun to sort of like riff and talk about all these tools. Yeah, it was more fun than I expected. I think if people like it, I think we should like go into like different categories because obviously with the AI video, there was like two other companies we could have listed. And like with AI art, there was one or two others. We could do like a... different categories and like actually spend time to like make sure we've tried the latest version of them too because some of these you know haven't tried them recently for sure well cool um i think on that note we should probably go ahead and wrap this one up uh thanks everybody for tuning in if you enjoy this type of content and you want to learn about the latest ai news and tools and keep your finger on the pulse and learn actionable strategies to use ai make sure you subscribe to us on youtube or follow us wherever you follow podcasts and Thank you so much for tuning in. Really, really appreciate you. And hopefully we'll see you in the next episode. Thank you.