I Tested 1000 AI Tools, These Are The 22 I Actually Use

So there's this one question that we get all the time. What AI tools do you and the AI Advantage team actually use in your day to day workflows? And I think it's a fantastic question because the actual use of something shows that there's not just value there, but that there's also a certain amount of stickiness there. As I like to put it, we try a lot of tools and use a lot of tools, but it's a different story to look at it three months later and evaluate which ones are still a part of our workflows. So that's why in this video we'll be looking at every single AI tool in our team's workflow flow that is AI powered. This includes some tools like Photoshop or notion that have AI features or LLM platforms like ChatGPT or Claude, and so many more in between. I was actually surprised by how many tools are on this list. And before we begin, we need to clarify two things. First of all, what is the work that we do so that you can contextualize all of these tools within our workload? At this point, the AI Advantage team consists of five full time people and another eight freelancers that have weekly recurring tasks. And without naming all the particular responsibilities and taking up another two, three minutes here the work is really divided into two categories. First of all, all the work on the YouTube and the free content that includes researching and testing all the releases, curating content for the YouTube videos and the newsletters and our tier lists in the free area of the community. This could be summarized that staying on top of the entire AI space when it comes to releases, innovations and functionality, and then us communicating that either through the YouTube channel, the newsletter, or the rankings in the free area of the community. And then there's the paid side of things, which breaks down into two subcategories. One of them would be the AI Advantage community, which is our learning hub. It's our school that teaches you practical skills and gives you personalized assistance. And here's where a lot of tools come in. Because over the past two years we built a lot of processes that rejuvenate the community. We create around 10 guides every week, hold around 20 events on various topics, and all of that is supported, supported by some of these tools and workflows that happen in the background. And besides some of the automations that we have, obviously the people are the ones that use all of these tools to create the value in forms of AI tool education. Okay, now that we covered that, I'm able to Go through this list as I'll be referencing some of the parts of the businesses when they relate to a specific tool or workflow that we use this tool in. And look every single week, me and the entire team made it our lifeline to try all the new tools and see if they can be implemented into our workflow. It's like literally what we do and although we try a lot, only a few of them stick around. So these are the tools that we actually use on a week to week basis. One very last thing before we begin and trust me, this is a fun one. We actually created something very unique over the past few months as a combination of many of these tools and also as a challenge to ourselves. It's a quiz that takes less than one minute and will recommend specific resources and prompts based on your interests. It's a more custom way of delivering free resources and and prompts right to you completely for free. We'll link it in the description and this will become important as some of these tools we only use in there because we really haven't found another meaningful use case for them. But in the context of this little quiz that recommends resources to you, they do make a ton of sense and they're a lot of fun concretely, the avatar generators. All right, so let's get into this. I'll just go alphabetically here, starting with the Adobe products and mainly we use Adobe Photoshop and Adobe Premiere Pro. Adobe Photoshop has some of the best AI integrations in the entire graphics design space. They have the Generative Fill tool. Now they even added a new AI powered selection tool that works super well. But as we do around 80% of our graphics inside of Photoshop, it's absolutely amazing not to have to round trip to other apps if you want to fill in a specific area if you want to select something. And as Photoshop was the very first software I learned back when I was 16 years old, that's 14 years ago. Even though I myself don't do this work inside of the company, I can appreciate how much of a productivity booster these AI powered tools are. In Photoshop it's incredible to just select over area and to be able to say what you want in there instead of having to mask it out, bring in another object and then making sure it blends with everything and then also having to artificially manipulate the light on there. The AI tools just do this in seconds and that way we can get more versions out in a shorter amount of time. It's just a feature none of us would want to give up anymore. Inside of Adobe Premiere Pro, on the other hand, we have this generative extend feature, which can come in handy, but in our productions, it's not super useful because we don't have a lot of B roll. And if we do have B roll, it is screen recordings, and there's no need to extend those generatively. If you would be doing something like event videography, like I used to do at the beginning of my career, this would be an absolute godsend. A lot of times you're missing a second of footage to cover something else up or to flow with the music. And there is generative extent tool would be amazing. Same with narrative storytelling, but we do tutorials, so it's a feature we use here and there to extend the clip to make things fit, but it's not essential. But what is essential inside of Premiere Pro is the ability for it to transcribe the entire sequence and to remove the gaps. This actually speeds up our workflow a ton. The first step in editing is always bringing in the footage, syncing it up and creating what we internally call the raw cut. It's taking out all the gaps of the footage and making sure all the pauses are removed. There's an automatic feature in Premiere that uses AI under the hood to transcribe the whole thing and to create this edit for you. We use this in every single project. And effectively, this feature alone saves us around five hours every single week, if not more. And that's on around two videos. If we do more, that number goes up from one feature. That's the power of AI Lovely to have around. And now let's move on to the next one, which is ArtFlow AI. And this is a platform to generate custom images. You can really easily train it on your face, and there's a lot of presets there. But now that Flux has come out, we started transitioning more and more to using those models, as they're arguably better in many situations, and as they're open source, you can also really easily train them on your face. We use Replicate for that. More on that later. But ArtFlow has a lot of these presets and it uses Dreambooth under the hood, which is just a different flavored image. And we still use it a lot of times because they have a lot of these presets where you don't need to be thinking up a new scenario or prompt engineering your way into something that looks good. They just have a plethora of presets where you can put your custom avatar into it and they work great. Next up, we have Canva AI and this is one that we experiment with on and off. As we have multiple Photoshop users on the team, this is our go to. But for things that need to be turned around quickly and a lot of social media graphics, like if we need a quick Instagram story story or even an Instagram carousel to communicate something, Canva is just fantastic because you have those presets there. And with the AI features it really does speed up the workflow. We don't really use the text generation features, but some of the image editing features and even some of the image generation features right inside of Canva can help speed up the workflow. Now, I think the Photoshop integrations with AI are way more significant and all the Canva AI stuff is all well and good. I think it's especially aimed at beginners. If you have more of a concrete vision and a lot of the assets predefined within the company as we do, and you have access and knowledge of Photoshop, a lot of it does become redundant. But when it comes to quickly producing a visual, Canva is fantastic and I can only recommend it. The AI makes it really beginner friendly. By the way, not a single one of these tools that I mentioned here are sponsored. I'm just giving you the raw take and we pay for all of these ourselves. Okay, next up we have chatgpt. I don't think I need to spend too much time on this. We talk about it a ton on this channel and it ranks as number one on rtlistrating. This is because it has the most tooling with the new model. It's also fantastic for creative writing, has a more human tone and things like that. Vast advanced voice Mode and the GPTs just set it apart from the rest of the pack. And the tooling really allows us to get the most out of this. We use it for everything from brainstorming or rewriting with O1 preview, all the way to data analysis and summarization and even more through their API. It's just our LLM of choice. And early on I made a decision to pay for a team plan. And every one of our team members is inside of our team, so we get to share GPTs. And it also makes sure that everyone in the team is at the forefront of what AI tooling makes possible as OpenAI really is pushing the boundaries of the space, which we're all about. Next up we have Claude and this is one that I have an individual subscription to and I know multiple team members decided to opt for an individual subscription. But really there's one thing that stands out with Claude and that is its ability to generate code. With Sonnet 3.5, it's absolutely incredible at that. And any development related project that we do, or any little script that I write for myself that I use in my workflows, I just go to Claude 3.5 sonnet and it gets the job done. I think everybody who's a bit deeper in the development department will agree with this. It's pretty much the consensus opinion at this point. Some people argue that O preview from ChatGPT is great and it is. It has longer output lengths, so that can be useful sometimes. But generally speaking, if you need code, just use Claude 3.5 Sonnet and you will be fine. Obviously all of these opinions are dated with the release of this video and the space evolves quickly, but as of right now, that would be my recommendation. I do also have to point out that we have a few projects going inside of Claude because we still really love the tone of it and before we used it for for all sorts of creative writing. But now with the new ChatGPT release it's become a bit more balanced. But we have a lot of projects where we uploaded multiple examples of our writing and we use it to draft up emails or events descriptions inside of our community. This is something that is pretty static and follows the same template every single time. So we just uploaded 10 examples of various event descriptions that have been written by hand and now Claude can help us craft new ones. We really like the tone of it for that. It cuts straight to the chase and sounds pretty human. Plus in an event description it doesn't matter as much as in something like an email or a video, which by the way, I'm speaking off the cuff here, I'm not following any script here. I have a database of all these tools that we maintain inside of the community. It also includes non AI tools that we use inside of our company. But really all I'm reading is the name of the tools and I just know how we use them because I'm involved in this seven days a week, 365 days a year. So next up we have Descript and this is a video editing platform that already existed before AI tools were around, but they really heavily leaned into this entire AI wave and I think they did a fantastic job. Matter of fact, from all the video editing tools that I've seen, I think they did the best job. They're trying to be AI native. That's at least how I perceive it they have all of these varying features like transcriptions and the removal of certain words, as Premiere does. But they go even further. They're repurposing into smaller clips, and you can do all sorts of edits on top of the transcript. So this means this tool allows you to edit a video as if it was a word document, not a video. You get to go through the thing and remove certain words and it edits them out. A lot of this can be done automatically with the power of AI. If you say, make it more concise, it looks at the entire transcript and it takes out the things that make it more lengthy. And we particularly use descript for any transcription related workflows. If we just want to transcribe something and get the timestamps, for example, of a live stream, well, we pull the live stream from YouTube, we throw it into Descript, and we hit the button that creates the timestamps, and there you go. Now, this could surely be done on a lower budget. I think descript costs around $30 per seat. I think in the company, we have two seats right now between me and Daniel, our digital content manager, but it is the easiest interface for this. Yes, we could spend a little bit of time and create an app that does the transcription and the timestamping for us, but we just appreciate the plethora of the tooling and they keep adding new ones. And as we're very, very video centric, we just keep finding the value in the script, although we do all of our editing inside of Premiere Pro. All right, next up, we have 11 labs, and this is the one where I need to reference the quiz that I pointed out at the beginning, because this is one of the main spots we use it at. If you fill out the quiz at the end, you will get a custom video that is generated with hey Jen. And I guess we can skip ahead and talk about hey Jen here too. And in hey Jin. I essentially trained up an avatar of me sitting here at this setup. And then we can use the hey Jen avatar in combination with the 11 Labs voice to look like me and sound like me, and we can generate any script that we want. Hey there. I'm glad you landed here. I'm AI Igor, and based on what you've told us about yourself, we recommend you check out the following resources to learn about AI. Now, generally speaking, I don't think the quality is as good as this, and that's why we don't really use it. I think there was one or two YouTube clips where we used it briefly. 11 labs we actually use in some YouTube clips. Sometimes if I get a word wrong or right before uploading, we realize there's some little inconsistency or some fact that I got wrong. We edit that out and regenerate the voice with 11 labs and it virtually sounds the same and I have to sit down and rerecord something from scratch. But beyond that, the main reason we use 11 labs and hey Gen in tandem is for this quiz where at the end you meet AI Igor for one minute and I give you the different resource recommendations. And the reason we went with this workflow is that as we update the resources or the prompts inside of the quiz results at the end, we can really easily edit the video. I don't need to sit down and rerecord eight new videos. Every time we make a new change, we can just slightly change the script, export the new clip, and there you go, we have a brand new video. So the speed of the turnaround, plus the different courses we're building inside of the community, we're starting to use hey Gen and eleven Labs as a part of it for me to explain certain concepts. And as AI keeps evolving, I don't have to sit down and re record something every single time a new update comes out. We can simply use these tools to regenerate a certain passage. And it looks, I don't know, 60, 70% of the way there to the real world recording. And when it comes to education, I think that is fine. The sound is flawless, by the way, it's the video that is about 60, 70% there. So there you go, that's 11 labs and Heygen and how we use it in our process. Let's move on to the next one. Okay, next up we have Flux and I want to discuss this one in combination with Replicate because that's the platform we use Flux on top of. You could totally run it locally, but I'd rather pay a little extra and not have it on my local hardware. Plus, other teammates can easily access the models we train. And this is the way we use fluxin. We train custom models with images of flow, mostly me, and then use these in our thumbnail designs. As you might have noticed on this channel, we find that the quality of the surroundings and the images overall is the highest. And the resemblance of me is pretty much equally as good as the other alternative, which is using stable diffusion on a platform like ArtFlow. So if you want to make custom images of somebody using Flux on a platform like Replicate or there's also file AI that I've used before is a great call and although you could run it on local hardware, there's no real point in that, to be honest. I've been looking for a reason, any reason really, to get a local AI computer. But the main reason that makes sense for is privacy for companies that want to run local models. And that's just not the case for us. We want to test out all of the cutting edge stuff and see how it performs. And in that case, renting GPUs is not just perfectly fine, but more cost efficient and easier. And you can make changes really easily if something in the setup doesn't work as you want it to. So that's flux. Next up on my list I have gamma AI and this is what Copilot365 in PowerPoint wants to be. It's an AI native presentation creation app and again none of these apps that we mentioned here are sponsored. But I'm just a big fan of Gamma AI because it makes creating presentations so easy. It works exactly as you would expect it. If you bring text in, there's an option for that. It maintains the text and creates a wonderful presentation from your text, or it generates all the text for you. It creates AI images, makes all the designs, and if you want to make changes, you can do things like go to a specific slide and say exactly what you would like to change in natural language and it will figure it out and do it for you. It works like a charm if you don't mind the watermark. The free version actually exports PowerPoint presentations with a little Gamma AI watermark on the bottom and you can just use that for school, university, whatever it might be. I really enjoy doing public speaking and teaching publicly and I learned over time that having a PowerPoint to support you really does help communicate your message. And Gamma AI is my go to when it comes to creating presentations with the power of AI today. Next up we have Gemini and I want to limit this one to Gemini inside of Google Meets because we use that for all meetings and actually having the Gemini assistant in there has been invaluable. Now there's many alternatives to this. If you use Zoom, they have their own built in tool and I believe most other video conferencing platforms also have something like this. But in our case we do use Google Workspace and Google Drive and Gmail and all that. And as a part of that we also use Google Meets. We find it really great. You can create these breakout rooms that has all the features we need and I pay for the Gemini upgrades there to have the meeting summaries works fantastically well. That's the way we use Gemini and we pretty much use that for every internal meeting. I will say that I have my executive assistant go over the transcripts and check if everything is correct just to make sure no hallucinations are present. It's a really quick process for her. If she's present in the meeting, which she always is, then it's not a hard task. It only takes a minute or two and it makes sure that our meeting summaries are perfect because sometimes hallucinations do sneak in. So you know, just be careful. That's the case with all LLM summaries. All right, next up we have make the automation platform and at this point I should put a disclaimer. We have been sponsored by them previously, but this is a platform we've been teaching and using before they even reached out to sponsor us. The reason we really like it is because the plans are very beginner friendly and you can export blueprints with prompt presets in them, whereas if you do that in some of the competitors like Zapier, it doesn't save the prompt inside of it. Plus you can get started on a free plan and just start building, start creating, start learning these tools without spending money, which is really great. Beyond that, these no code automation tools are just a no brainer for us. We use them in so many ways. We have a bunch of automations for our community, so in notion we maintain a bunch of databases with all community members and for example the info they fill out in the onboarding form. And then some of these make automations help us use something like a fine tuned model that we create to classify people into different districts by interest. So if they come in and they express acute interest in creating photos and videos with AI, then we sort them into the creative district and if they want to learn more about automations we sort them into the no codes district which has content specific to that. Now this is really easily possible and editable and updatable through a no code automation tool like make. So we have a bunch of automations running on the background. I think in total in the advantage we have around 35 to 40 automations running right now. Some of them very simple, some of them more intricate. Like hey, if person has outcome number six in the quiz that recommends the various resources, then make sure to fire a webhook which will be picked up by the automation software and then we can tag them inside of our email marketing software and send you a custom Email with your resources. Plus, in the future, if you do decide to sign up to the community, we can help you get to the relevant resources faster because in the past you already expressed interest in a specific topic. So we can recommend materials that go along with that. To be fair, some of this stuff is more advanced, but hey, tools like this make it very accessible as opposed to doing everything with custom code where it does become a bit trickier to maintain. Maintain. Okay, next up we have Midjourney. This is sort of our go to for all the art. We have a lot of banners and graphics inside of the community and also the shop now, but mainly in the community for all the events and all the guides and resources. We create custom banners. And Midjourney is our go to tool because it has the most appealing visuals. We created specific styles and have a specific visual language that we follow there. And by now it's really easy to maintain a consistent style inside of midjourney. So we do that and then we can just generate new and new visuals that that are relevant to whatever the content is. So midjourney still best in class in terms of art, in my opinion. Next up we have NotebookLM. And this one is sort of a no brainer at this point. Everybody loves NotebookLM. If you're learning something, if you're researching something, this is your go to tool. You put all the sources in there and then you can generate these podcasts to either refresh your mind or retain the knowledge or even communicate it to somebody else. It's just a great way to do it. This natural podcast is just a thing that worked and I wouldn't want to go without it anymore. I don't use it all the time, but when I'm learning something, you're exploring a new topic, it's my go to. I put the sources in there and it really helps in the process of exploring a new topic. Okay, then we have Notion. I'll keep this brief, but basically become sort of notion obsessed that everything we do is inside of Notion. From all of the research workflows, to all of our video production workflows, to all of the community management. Everything from cataloging new ideas or feedback all the way to managing the guide flow or even the event scheduling with external speakers that we have inside of the community. Everything happens in Notion. I just wanted one platform where I can build it all. And the wonderful thing is it integrates very well with some of these no code automation tools, meaning we can really easily plug it together with something like the ChatGPT API and then work with the data inside of the notion and alter the databases, extend them, whatever we want to do. Everything is in notion and I really like their AI features and I paid for the add on. I don't use it all the time, but sometimes it's just a fantastic way to manipulate tables into databases, into articles, into to whatever you might need. I really use it for content transformation as everything lives in notion and then the AI helps me transfer it from one format into another. Then we have Perplexity AI, my Go to AI search engine. So I'm still kind of flip flopping between ChatGPT Search and Perplexity AI right now. They're both great. I don't think they replace Google at this point, but when it comes to something like product reviews or any human opinion, I will always go to Perplexity AI. Also when I have factual questions where I just want the answer or maybe I want a little bit of research and info, then I also like Perplexity AI and actually my favorite thing is you can ask concrete questions and compare two things and it will look into that. So hey, should I buy this headset or this headset? And then it will come up with an answer for you with the various resources supporting that. If you haven't tried it, I would recommend it. There's a reason that we consider this a S tier tool and everybody seems to agree. Then we have Runway and we kind of play with this. We really love this tool in the video production. We use it a lot to remove the backgrounds of videos. So basically green screening something and then also more and more we use it for B roll where we generate an image with midjourney or flux and then we animate it with Runway and we turn it into a video which we can then use as a B roll clip inside of our YouTube videos. It's a great way to freshen things up. And something that on a generative AI educational channel just makes sense. Other than that you can use it to create transitions, but I don't know. This is sort of in the experimental category. The killer use cases haven't surfaced with these tools yet in my opinion, but we keep using them to stay on top of it. I feel like once we get character consistency, it's going to be able to do a lot, but that's just not the case right now. Nevertheless, a powerful tool that we keep playing and experimenting with and our Go to Video generator because of all the tooling. And then lastly on my list here we have Zapier, another no code automation tool. This is the one I was used to from back in the day from my other video production business. I already used AP in the background. Now this is probably the most expensive no code automation tool. And the reason is because it's the simplest, they kind of set it up in a way where it's quite intuitive to get started. But with that being said also make is quite intuitive to get started with all the tutorials. We do prefer to teach inside of make because you can export the blueprints so easily and they contain the prompts and the pricing is more accessible than Zapier. But I'll just say that there is specific connections and apps that just work better in Zapier, others work better in make, but as you kind of move through this app, sometimes something doesn't work in make and then you try in Zapier and it works perfectly. And the other way around, I kind of like having both. That's just me making a living off of experimenting a bunch in these tools and also teaching them to others. And that is essentially it. Those are all the tools that were really sticky that we keep using, that we keep coming back to. If you're interested in learning about what's new and what you could add to this stack, you can check out the Friday show news you can use where we cover all the new AI use cases of that given week. And then in a year or something I'll update this video. As mentioned in the community, we update this regularly and also it includes all the non AI tools that we use in our tech stack. But other than that, I hope this was helpful. If I missed any tools that you feel like belong on here, the comment section is there exactly for that. I would love to hear which AI tools you use regularly that might not be on my list. And yeah, with that being said, I hope you have a wonderful day.