Matt Wolfe Ranks The Best AI Tools For Marketers In 2024

On today's show, we got a hot one. We are doing an AI power ranking. We're going to tell you the hottest and most important AI apps to use in your marketing today. We're going to be breaking down all the use cases and we're going to be doing it with a very special guest. Kieran's out today, but I am joined by the one and only Matt Wolf. We're going to tell you the best apps for translation. We're going to tell you the best apps for text to video, text to image. So much coming. This is a jam packed episode. You're gonna wanna stick around to the very end cause Matt also has a special announcement. Let's get into today's episode. We'll be right back to today's show. But first I gotta tell you about.

A week that you're not gonna wanna miss. It is a highlight of my year every year. It is Inbound, which is HubSpot's annual event where we bring together the best attendees, the best speakers, the best experts in sales in marketing and customer success in AI. We all learn together and we all figure out how to grow our businesses better in the coming year. And this year it's no exception. It's a fire lineup. It's going to be at the Moscone center in San Francisco, California, September 3rd through the 5th. So it's a great way to kind of close out Q3 and kick off Q4 really strong and have your heads right for 2026 because you're going to learn and hear from Amy Poehler, Dario Amadi of Anthropic, Dharkesh Patel, Sean Evans from Hot Ones, Marquise Brownlee, Glennon Doyle, and the world's best speakers and experts across marketing, sales, customer service, AI. They are all at Inbound. Hundreds of amazing sessions. And you know, it's more than just the sessions. It's about the networking. It's about learning from your fellow attendees, your peers, understanding what people are doing really well, where people are getting stuck. That is how you break out. That is how you are going to grow better and faster in 2026. You're not going to want to miss it. You can go to inbound.com register to secure your ticket and your spot for Inbound 2025.

Matt, welcome back to Marketing Instagrain. This is your second appearance and I think it officially makes you a friend of the pod. So thanks so much for joining us.

Yeah, happy to be here. Always love nerding out about tech and AI. So it's going to be Fun.

And as you may have heard in the intro, you're also for sure going to want to check out Matt's new podcast called the Next Wave. And you can check it out at the Next wave pod on YouTube. You should definitely go check it out. Hit that up on YouTube. Subscribe if you. It's going to be awesome. Matt, you and I are having fun today. Kieran is off as a new dad and we're going to basically play on an episode Kieran and I did. About a year ago. We did a marketing channel power ranking. We did a power ranking of the channels marketers needed to use in the moment. And now we're doing AI tools that marketers need to use in the moment. And I wanted to start, Matt with honorable mentions. Anything that we're like, oh, maybe it's not in the power ranking now, but six months from now, 12 months from now, maybe it could be. And I'll give you mine first for you to react to, which is. Have you seen. I think it's Hume. H U M E. Yes. What's your take on Hume?

Yeah, so Hume is really interesting cause Hume actually listens to like the inflections in your voice. It listens to try to figure out if you're angry or happy or sad or depressed. And it notices like micro inflections in your voice to add additional context to the discussion you're having with it, which I think is going to be really powerful down the line. But it is still very, very early. I actually played around with Hume a little bit on one of my videos and I just talked to it normal, like the voice I'm talking to right now. And it said, I detect anger in your voice. What's wrong today, Matt? And I'm like, where did that come from? So it's, it's got a little ways to go, but I think it's really, really interesting.

Yeah. And I think why it makes my honorable mention list for marketers is because the Hume demo that we just showed a few minutes of and Darren will link up in the show notes. What's interesting about it is the demo is essentially like a real time concierge on a website, which you could imagine if you could talk to a website, it can detect your emotion, get the details of what you're trying to get through, not just from the specific words you're using, but from everything about your voice. That being a very powerful way to guide a website. Make sure you're finding the answers you need as quickly as possible. Which is exactly what a marketer is trying to do. It's not to a place where you can do that today, but I wouldn't be surprised 6 to 12 months from now if we're talking about this as a use case of PUM or some other similar technology as having kind of these guided voice concierges for your website for marketers, which I think is pretty freaking cool.

Yeah, it's interesting too, because when I was using Hume, I asked it, can you tell me more about Hume? What is it good for? And it actually was jumping me around on the site to different sections of the site as I asked questions. So that was really cool as well. So I can definitely see it being used for like a customer support kind of thing. Hey, I have a question about X. Yeah, it knows the question and it puts you to the right support document, but it's also understanding, hey, is he upset? Is he happy? Is he like. It's also taking those clues and I believe it's also going to watch your face and detect facial expressions down the line as well and use that as additional context. So definitely. Really, really interesting tech that I'm keeping my eye on for sure.

All right. Anything else in that? Like, hey, I'm keeping my eye on it because it's not quite ready for primetime, but the promise of it is very interesting and we should put it on our honorable mention list before we get into kind of the apps that are more ready for primetime today.

So one tool that comes to mind that I actually think is really, really good already. Right. It's kind of already there is Suno. I don't know if you've had a chance to play with Suno yet.

No, I haven't. Gimme the rundown.

So Suno is an app that creates songs. So you give it a prompt and you say, oh, yes, yes, I want a song.

I've seen the demo. I haven't used it yet.

Oh, it's so good. So you can say I want a song in the style of a early 2000s pop punk anthem all about why people should listen to marketing against the grain. And it'll make a whole two minute song about marketing against the grain in that style of music with lyrics with background music. And the song is really, really, really good. The only reason I would put it in the honorable mention section is I feel like it's more of kind of a novelty, like a fun thing to play with yet, you know, unless you need like a jingle for your business or brand, there's not really a whole lot of use cases for a business yet. But it is probably one of the most fun tools I've played with in the AI far.

Well, what's interesting is I imagine for a lot of people out there, like if you're at a large organization, you're running high volume audio or video ads, you could have a lot of different jingles that you create very quickly and test. Right. Like some type of persuasive song or you could create some type of highly different video creative that is essentially like this AI song that kind of sets you out. So like there are real use cases there. They're just not going to be common and they're not going to be for everybody, which is why it's not kind of at the top, top of our list. But I think that's a great way to call out. I've heard a ton of people being very impressed with what they've been able to make with Suno in the early days, which is really cool.

Yeah. And it can do full length verse, chorus versus chorus kind of songs and yeah, I mean as somebody who has a YouTube channel, it's great for making like an outro music.

Yes.

You know, song for my video or something like that. So there's use cases on playing with it. But if you, if you're like a small medium business and you're like suno is fun to play with but you may not actually see a need for it in your business yet.

Totally agree with that. Okay, so couple honorable mentions. Now let's get into the real deal. We want to share a few AI apps today that we do think are ready for primetime. And the question I normally hear from people, Matt is like, where do I start? Like I know there's this AI stuff out there and I'm a marketer and I'm just trying to get like my emails out. I'm just trying to do the basics stuff I'm trying to do. Can you give me something that I can just like go and start and do? And what we're about to tell you, these are going to be things in that category where you can go and actually get started. And there's a real use case and I think definitely one of my tops is Heygin and I know it's one of your faves as well. I've spent a few sessions with Josh, the CEO over and founder over at Heygin and one of the reasons I love Heygin is because they innovate quickly. If you're watching today's show like a week ago, week and a half ago, probably when this airs, hey, Jin went completely viral because they have figured out a way to create a video avatar that can walk and make real emotions and hand gestures. And it's like. It is incredible, Matt.

It is really good. Yeah. I mean, the translation, so you can translate your videos into any language that you want with it, and it will actually still use your voice and. And it'll actually make the lips look like you're saying the things you're saying. In this new language, you can do bulk video creation. So let's say you're emailing a list of potential clients, and you want all of them to see the same video, but you want to inject their name and maybe their company name into every single video. You can, at scale, create, you know, 20, 30 videos that all say the same thing, but then inject the person's name and company name into the video so it looks like it was personalized to them. So you can do things like that, that make it really cool and. Yeah, hey, Gen in Motion, I think, is what they're calling it. That's the newest feature that they rolled out where you could record a video of yourself walking down the street, maybe just saying, blah, blah, blah, blah, blah, as you're walking down the street. You don't have to really be saying anything, and then go and dub it over with whatever you actually want coming out of the mouth of the person walking down the street.

Yeah, it really fits a lot of that. Like, TikTok reels kind of creator influencer use cases where you're out in, like, full motion and being able to create videos in that style, which is super cool. On that note, on translations, what I would share with you, Matt, is like, we did a test internally at HubSpot, and we found that the Heygin translations, in terms of engagement, view duration, and everything on YouTube, perform just as well as if you did a full localization with a local language speaker. We did it in Spanish for our test, but it's good. Yeah, it's not perfect, but it is so good that it's above that threshold that people will watch, that people will engage. The fact that it's in your own voice and the lips are dubbed pretty close to perfect. There is very, very important. And if you are a multilingual business, you should pause this podcast and start going and doing some work with hey, Jin. Because there's a real opportunity there.

Yeah, absolutely. It struggles with beards a little bit. If you've got a beard, sometimes the, you know, the mouth, the beard doesn't look like it moves with the Mouth properly, but it's really good. Couldn't recommend it enough. I've spent a lot of time playing with it. I've done videos on hey Gen where I've had it translate to Spanish, Japanese, German, French. And I actually asked my audience, let me know in the comments, does this translate well? And for the most part, everybody said yes. That looks good. Maybe it's a little bit more formal language than we'd use in real life. Maybe there'd be a little more slang or a little bit more sloppy speaking. And this is like trying to polish it up and be very formal, but other than that, it gets the point across.

Yeah. So number one on our power rankings for this show. Hey, Jin. All right, Matt, now that we've moved on from hey, Jin, what is another one that you would put up for, like, a marketer has to go and explore and use this today.

Now, this might be cliche. I don't really know, but for me, I use Claude every day. It's bookmarked. It's like, open in one of my tabs constantly. For me, it's replaced ChatGPT. I know probably every single person listening to this has probably at least heard of ChatGPT or use ChatGPT. But now Claude, I think, is even better than ChatGPT. And pretty much all of the benchmark testing, it's done better. The context window is larger, meaning the amount of data you can feed it and the amount of data that it can feed you back is much larger than any other publicly available model. Right now, I use it for summarizing articles. I use it for turning things into bullet points. I use it for just brainstorming ideas for videos. I use it for brainstorming ideas for thumbnails for my YouTube channel. I am in Claude constantly, so I feel like if everybody's worldview of AI is chatgpt, maybe test this new one and go play around with Claude a little bit.

Well, first of all, it's a sad thing Kieran's not here, Matt, because he and you would be battling for who is the number one fan of Claude 3? Because Kieran Flanagan loves Claude 3. He freaking loves it. And he should, because it's awesome. I love it, too. But he is like an unabashed Claude 3 fan. So I think what we're saying here is for that kind of marketing copilot, that copilot that you're going to maybe ask to edit some work, maybe help do some research, maybe say, hey, I'm working on this product positioning, what are some ideas that I'm not considering. What are some ideas for this, for product naming that I might be doing? All of these types of use cases, we think Claude 3 is the best suited for this right now. And from my own use, Kieran's use, it seems like from your use that's really the case as of right now. Doesn't mean that that won't change. But right here and now, Claude 3 is the best.

Yeah, that's my opinion. There's only two things that I feel like ChatGPT actually does better right now. ChatGPT can actually search the web. Claude still doesn't search the web for you. So that's sort of one of the downsides of Claude and the other is that ChatGPT has Dall E built into it. So if you want to generate images directly inside of it, you can do it inside of ChatGPT. Saying that we have Bing Image Creator, which uses Dall E3 does the exact same thing that ChatGPT does with images and that's free for anybody to use. So you're not really missing out on much.

Right. But if you're just looking for a text based copilot to go back and forth with your marketing work, while ChatGPT is also awesome, Claude is going to be a little bit better. We found the quality and the right thing with these copilot use cases really does come down to what you said, Matt. That context window, how much context can you give it to get the best Response back? And Claude 3, the context window is so huge now.

Yeah. And you know, to give more context, a 200,000 context window is essentially 150,000 words. So the rule of thumb is about 75% of the context window is the amount of words input and output. So 200,000 context window means that you can upload a book that has 150,000 words and it will actually be able to find information from within that book.

That's pretty awesome. So, hey Jin, Claude 3, where are we going to go next? Matt, There are a bunch of great tools that we're using. One that we're using for this pod Opus clips. I would probably put Opus clips on our list. You know, we've got a long list of stuff we're using, but Opus Clips helps and cuts short form video clips out of long form videos. It's exceptionally helpful because the way the Internet works today, you normally, if you have a long form video, you want some short form videos out there to go and kind of drive people back to it. Have you been using Opus Clips. What's your take on Opus?

Yeah, I've been using Opus Clip. You know, one thing Opus does is it uses its AI model to try to hunt down the spots that it thinks has potential to kind of go viral and then clip those out into like a minute segment. It's hit or miss for me. I actually really love what it does, but it'll generate, you know, 10 clips from one video. And out of those 10 clips, I typically will find one or two of them to actually, you know, be good enough that I'd actually want to share it around. You know, sometimes it might begin a conversation and then cut off the conversation before it's over. So you don't, you know, you don't get the punchline essentially, but it's pretty dang good. It's the best one out there that does what it does.

Yeah, completely. And you know, if you're pressed for time, you don't have an editor or you're doing something part time, you know, video is not your full time job. The time saving trade off is going to be worth that because you're still going to get one to three kind of clips that you can use.

Yeah, and I'm actually really impressed by how well it edits those clips as well. Because if you, you're on camera speaking and maybe you're also sharing your screen, it will actually jump back and forth to making your face take up the frame and then jumping over to making the screen take up the frame. And then it also adds the subtitles like you see in all the TikToks and YouTube shorts as well. So it actually edits pretty dang well to a point where you're like, okay, I can actually see this doing well on TikTok or YouTube shorts or reels or whatever. It does a really, really good job at that. I think where they just need to hone it in a little bit is, you know, hone in that AI on finding the exact perfect clip. Because while it is really good right now, it is kind of hit or miss and like maybe 20% of the clips it creates for you will actually be usable.

Yeah. And it kind of falls in this category that is kind of getting defined as content remixing. Right. And a quick plug for HubSpot Content Hub, which literally just got launched. I know you haven't even tried it yet because it launched yesterday. And there's a great AI app as part of it's called Content Remix and takes one of your content assets. Maybe it's a blog post, maybe it's A long form video. You know, if it's a blog post, maybe it's going to use 11 labs and help you cut it into a small audio file for a podcast or an internal update. Or it's going to take this video and do video clips. It's going to do that same principle, which is regardless of what tool you use, you're going to want to do some type of remixing to say, I've got something that I spend a lot of time and effort on that I think is good. How do I merchandise it and get more out of it? Which is really what we're talking about. Opus clips, content hub. There's going to be others that we'll I'm sure talk about on future episodes. But like that category, if you're a marketer, is becoming way more important.

Yeah. And you know, another tool that you just mentioned is 11 labs. I would put that in my Mount Rushmore of current cool AI tools as well.

Yes. Let's just get into it. It's definitely on the power rankings. 11 labs. What's interesting is that all these apps launched within like four months of each other. Hey, Gen ChatGPT, ElevenLabs, Runway. All these things launched literally within four months of each other. We've talked a little bit about elevenlabs before on the show, but please give everybody the rundown on 11 labs. What you think the killer use cases for 11 labs are.

Yeah, sure. So 11 labs is a text to speech generator. Uses AI where you can enter any sort of text and it'll speak it out in a very realistic voice. We're way past the days of the.

Old Apple computers where amazing like you cannot distinguish.

Yeah. You cannot tell it's not a real human speaking for the most part. Right. It could be a little monotone at times where you can get hints of it, but for the most part it's pretty good. And it can actually learn any voice, I think, with 30 seconds of sample size. So I can upload a YouTube video of me speaking for a minute and it can learn my voice and then generate audio in my voice. They also have a voice marketplace, so if you don't want it to be your voice, or maybe, you know, you have a strong accent or you just feel uncomfortable speaking on camera. You can use one of the voices inside of their marketplace as well and have a realistic human sounding voice. But I see all sorts of use cases for this. You know, as a content creator, if you're doing podcasts or YouTube videos and you need to do a quick overdub in the quickest way possible. You can have 11 labs, you know, say a sentence for you, dub it in, and quickly edit that. But I think the even bigger use case for more businesses is to create multimodality with your content. Right. If you have a blog post, if you have news articles on your website, things like that, plug them into 11 labs, get the MP3, and then put a little audio player at the top of the blog post so people can press to listen to the blog post instead of having to read it. And, you know, for accessibility reasons, you know, some people have a hard time reading, the audio is better for them. If somebody's driving and they're like, oh, this is a really good article. I want to listen to this article instead of read it. Well, they can open up your webpage, press play on the button, listen to it through their Bluetooth speakers in their car. Right. So I think it's really, really good for remixing your existing content into an audio version of the content.

Yeah. What we would say is, if you ever need audio and you have the information, there's no excuse to not have the audio anymore. Right. Like, that's really what you're saying. It's not cost prohibitive, it's not quality prohibitive, and it's not even time prohibitive because it's super quick.

Yeah. And they've got APIs too. So it's really easy if you know what you're doing with things like zepia or make.com or one of those kinds of, like, integration tools. You can build little workflows where you're not Even logging into 11 labs if you don't want to.

No, it's just automatically going to create it in the background and give you that file every time you publish a new blog article or a new newsroom article, whatever it may be, Right?

Absolutely. And they have translation features where it will actually translate whatever you put in into whatever language you want and it'll still sound like your voice. So it's like what heygen does for video. It does this specifically for audio. So I can plug in a whole blog post, tell it to translate it to Spanish, and then I can, if I wanted, have a second version of that blog post in audio form in Spanish. They also just released a new feature for sound effects. So you can go in there and say, I need a sound of a siren or a dog barking or, you know, somebody dropping a box of soda cans or something. Right. And it will go and generate those sound effects for you too. Probably a little bit less Useful for like small businesses. But if you're a content creator, it's actually pretty dang useful.

And one of the themes of today's show is that one of the early use cases, if you're a marketer, is that translation. Multilingual use case. Right. Because Both Heygin and ElevenLabs make it so easy for you to get really good translation very quickly and very cheaply, which is especially if you're a small, medium sized business, that has been a real impossibility up until really like the last six to 12 months.

Right.

And that is a game changer. And so sometimes remixing your content does just mean a different language. It's not anything more sophisticated than like, hey, I did this in English, it'd be great to have it in Spanish and French and German. Right. Like we can do that. That is like within our control to make happen.

Yeah. In every modality. Video, audio, text. It's pretty easy no matter what sort of form of content you're putting on the text side.

What do you think is best on that text translation side? Right now, since we're, since we're having it.

So I like a tool called Deep.

L. Sorry, spell that again. Cause I didn't quite get it.

Deep L just D E E P. Wait, D E E P L dot com.

Okay. Deep L. All right.

Yeah, so it's a translation tool that uses AI to do the translation. You know, there is Google Translate, which is really easy to access. You can, you know, copy and paste right into it. It'll do it pretty well. But DeepL, I believe has like more of the AI neural network built into it, so it actually works a little bit better. But yeah, there's a lot of stuff coming out that can do that. I don't know if Descript does translation yet, but if they don't, I'm sure that's coming. And tools like descript. But right now for text translation, DBEL is the one that I go to.

Hey everyone, if you're like me, you love learning from the best and you just love learning, period. And a podcast that is all about learning is one that I love. It's called Billion Dollar Moves. It's hosted by Sarah Chen Spellings. It's brought to you by HubSpot Media. And Sarah's awesome. And she covers really what it takes to build a remarkable business. She's a venture capitalist, she's a strategist, she has really hard questions and wants to find the highs and the lows from the most successful builders and entrepreneurs of our time. So that we can all learn from them and make our own billion dollar moves. She's got a great episode around the Big Bet mindset and how you really build bold change for your company. It's awesome. You're going to want to check that out as well as all of her episodes are awesome. Highly recommend. You can listen to Billion Dollar Moves wherever you get your podcasts.

All right, so Deep l for text, 11 labs for audio. Hey, Jen, for video, you got your Mount Rushmore of translation for any of your marketing content right there, which is pretty incredible. And something that is also. None of those are super expensive or cost prohibitive to really do. Right.

Most of them even have free trials, free plans to play with when you upgrade 11 labs. 11 labs is 5 bucks a month to upgrade. And, like, it's crazy to me how inexpensive it is to be able to use 11 labs and tools like that.

Okay, next we're gonna start with the use case. Then we're gonna give an app or two. Because I think one of the best marketing use cases right now is text to image. And the reason for that is if you're a marketer, you've hated using stock photography, you've got that perfect ad copy, but you don't have the right image to go with it. You've written that perfect article, but the image doesn't quite go. It's one of the most frustrating things in the world is to find the right image with the story you're trying to tell. And text to Image does an incredible job of solving that.

Yeah.

And I think we should give folks the top three text to image. We think right now to help people kind of go and get started. Like, for example, we used Midjourney, which is a great text image product, just for, like, ad images for, like, Facebook ads. And we saw a massive improvement because you could have the perfect ad image to go with your ad copy.

Right.

It seems so simple, but that's just an easy win. You can go and do.

Yeah, for sure. And pretty much every single thumbnail that I use on my YouTube channel now is an image I generated with one of these AI image generators. But it's funny you said three, because there's three that I sort of rotate between depending on what I need them for midjourney being one of them. I love, love, like, the color palettes and the contrast of midjourney. They just. There's something really pleasing to the eye about the images that midjourney generates.

I totally agree with you.

There's like a really, like, artistic style where you can see the image and you can kind of go, I can tell that's mid journey. It's got that mid journey look to it and it's really, really nice looking. But Midjourney also has some issues with prompt adherence, where if you give it a sort of complex prompt of I want somebody on a hoverboard with a lake behind them at night, with the moon shining behind them and a cityscape, some of those elements are going to get missed, right? If you want something that's really, really detailed and you want to put a lot of things into that prompt and you want it to make sure that that image has all of those little elements that you put into your prompt. There's nothing better than Dall E3, which you can use through ChatGPT. You can use it through Bing Image Creator. In fact, I think if you just go to like Bing Chat or I guess it's called Copilot now, if you just use Microsoft's Copilot, you could generate images directly with Dall? E3 right now. And they're super prompt adherent, really, really good. Like I did a test where I said I wanted a three headed dragon sitting on a couch, wearing a fedora, eating nachos and watching MTV or something like that. And it got all of the elements in that image. It didn't miss a single one. I don't think that the images Dall E generates are as aesthetically pleasing as what Midjourney generates. But it's really good at getting all of those elements into the image. And then the third one is stable diffusion. It's an open source model that's completely uncensored. You can make anything you want with it. Literally nothing is off limits to what you can create with stable diffusion. But stable diffusion is very, very complicated. You need to install it on your computer. You need to use like a user interface, automatic 1111. And you gotta, you know, be comfortable working with your terminal on your computer. It really only works well if you have like an Nvidia GPU in your computer. It really only works well if you're on a PC and not as well on a Mac. You know, there's a lot of downsides to using stable diffusion because of the complexities, which is why I like a tool called Leonardo AI, which is stable diffusion under the hood. But Leonardo is sort of that front end user interface that's in the cloud that can do everything stable diffusion does, but you know, from a web interface and they also have like a credit system where every day you get, like, 150 free credits a day. So most people can just use the free plan and just generate whatever images they want. And again, totally uncensored, you can do what's called control nets inside of it, where you upload a specific image and then you tell it to generate an image that matches the, like, pose of that image. So if you have a person who's, you know, going like this and doing, like, some sort of shocked face or something, you could find an image of somebody online doing that face, pull it into Leonardo and say, I want this same pose, but make it a different character. And it'll generate an image that follows that pose using the control. Net and give you that image that looks exactly like what you want. So the most flexibility, customizability, Leonardo AI the most prompt adherent if you want all of these elements in your image. Dall E3, if you want that sort of midjourney, beautiful aesthetic. Also realism. Midjourney is probably the best right now at getting something, like, really realistic. Those are the three tools that I use and what I use them for.

I'll give a bonus out there. Have you used Recraft yet?

I have not used Recraft.

All right.

Check out Recraft AI. I think very similar quality output to Midjourney, but they give you a vector.

Ah, nice.

So if you are a more advanced user, you're a marketer who's used to using Photoshop, working with vector images as part of a bigger graphic design work, you're doing Recraft, that is gonna be your friend because they're not just giving you a plain image format like a JPEG or bitmap or whatever. They're going to give you a vector image which is really, really differentiated in that space.

Yeah. Great for logos, too. Right. If you need to generate a logo, you want it in a vector image because you want to be able to infinitely scale it up. You want to be able to pull it to Illustrator and maybe put your own text over it. Another one that I've played around with a little bit is Ideogram. Ideogram is probably the best in the market for adding text to your image. So if you want to do a prompt of, like, a wolf holding up a sign saying subscribe or something like that, it will actually generate that image and the text will be legible. It'll actually be the right word. So ideogram is kind of leading the charge with image generation with actual legible text in it.

Well, we said three. We gave you five. I'm sorry. Sorry for the bonus look, text to image is I think, probably the most important marketing use case of AI today because it's just kind of universal to so many aspects of marketing. Whether you're a solo marketer or you're a marketer at a big company, you're going to use text image and you're going to use it pretty regularly. And so I think it was really worth kind of diving in and doing that and kind of making sure we were covered there. All right. This has been a pretty thorough rundown of our current marketing power rankings in terms of these are the AI apps you should really care about. Do you think we left anything out, Matt? Is there anything out there that if you're a marketer, it's like, no, you got to go check that out too.

I think the only thing we didn't really touch on is the text to video. Right. I mean there's hey gen, but hey gen is kind of using your avatar, your face and then you know you're putting words into the mouth of that avatar. Right. But there's the other side of the text to video where you give it a prompt and it just, just like the AI image generators, it's generating a video of what you prompted.

Yeah. So for the marketers out there, you would call that a roll versus B roll. Producer Darren's happy looking. Producer Darren's like, whoa, look at him talking about some video terms. And hey, Jin is a great a ro B roll tool. Right. A person looking at a camera being able to describe something where text to video. Most of the other ones are more B roll tools. Like, we've seen SORA gain a lot of popularity as a text to video tool even though it hasn't been released. OpenAI keeps teasing the quality of videos that have been made and they've been releasing some kind of mini movies and stuff. So then that begs the question, Matt, what is your favorite kind of B roll text to video app that people can go and use today?

So there's three that I'll mention real quick. Pika Labs and Runway. Those are two that kind of do similar things. You give it a prompt, it will generate a video based on that prompt. The videos are usually three to four seconds and then they have a button to extend it. And I think you can extend it up to like 16 seconds or so. So you're not going to get super long videos from it. But like you mentioned, B roll, B roll, you typically only use a 4 or 5 second clip anyway for B roll, so it's fine for that Use case. They also have like inpainting where I can generate a video and then sort of select an area of the video and go, all right, I want to change just this section of the video, which is really cool. That's kind of one of the newer features we're starting to see in some of these AI video generators. The new one that's, that's starting to roll out is one called LTX Studio from a company called Lytrix. And LTX Studio is one where you can give it a prompt and it will make a minute, two minute video for you. But it uses similar technology to Runway and Pica, where it's generating three, four second clips, but it's stringing a whole bunch of them along. It gets you a long video based on your prompt. And then you can actually get in there, see the storyboard of the whole video and say, all right, the shot that it used in, you know, the fifth shot in this video isn't quite right. And then you can go in and reprompt each little clip individually if you need to. So it's kind of like doing what Runway and Pika do, but at a more like bulk level where it will generate the entire video footage. Yeah.

If you're doing maybe a long form or you needed some abnormally long B roll, like, that's going to be really compelling. Where if you're doing a video with a lot of cuts and you need like the very specific short bursts of B roll, maybe something like Runway or Pika is going to be a little bit better in those use cases.

Yeah, yeah. And I mean, I'm sure there's some decent use cases for video ads on Facebook and places that I don't do a lot of paid media these days. But, like, you know, I feel like it'd be really fun to play around with some of those kinds of tools for video ads on places like that.

I completely agree. I cannot believe how much we crammed today's show. This is like, you know, I was thinking before we recorded today, Matt, I was like, who does Matt remind me of? And you know what you remind me of? You remind me of Mr. Wizard, but for A.I.

Oh, yeah.

Do you remember Mr. Wizard? I love Mr. Wizard so much. It was a show on Nickelodeon where he taught everybo.

Okay.

Yeah, yeah. He would, like, do the experiment where, like, you stand on your head, like against the refrigerator and swallow an apple upside down to show that you can swallow upside down, like the vinegar volcano. All those things where it's like, take this thing that People don't really understand and explain it in a really good and simple and usable way. And that's what I felt like you did today. And I really, really appreciated your help in doing that because I think we just gave all the marketers listening some things that they can go do today to start using AI, that's gonna help their marketing. That's not gonna be overwhelming. It's not gonna cost a fortune. All of those things are gonna be really important.

Yeah.

All right. But before we go, Matt, you are doing a new show on the HubSpot podcast network. It's called the Next Wave. It is all AI. If folks love today's show, they're going to love the new pod. Tell us all about.

Yeah. So this podcast has been a blast. We're working with producer Darren over here to put it out. And it has been such a cool experience working with HubSpot to build this show. But the concept of the show is that me and my co host, Nathan Lands, who is, you know, very deep in the AI space as well, we have conversations every week. Our goal is to be your business's chief AI officer, keep you in the loop with the latest news in the AI world, keep you up to date with the latest tools. Very similar stuff to what we talked about today. We're bringing on amazing, fascinating guests that are building things in the AI world. So a lot, a lot of amazing conversations that are going to keep your finger on the pulse of what's going on in AI. And we're launching really, really soon. So check it out wherever you listen to podcasts on Apple, podcasts, Spotify, everywhere else. We're also doing video versions over on YouTube, so you can find it over at the Next Wave pod over on YouTube. But super excited to get this out into the world.

Well, I know I'm going to be subscriber to Next Wave Pod. I hope everybody listening to today's show is Matt. Thank you so much for joining us. Can't wait to have you back on again in the future.

Thanks so much for having me. It's been a blast.

We'll see you real soon on market. Against the grain.

Sam.