So you're referring to which, by the way, those three accounts do about 200 million views a year and they're all fake AI personas with a little comical twist, but we make an effort to be as authentic as possible when we're replying to comments, DM, ING people, etcetera.

So Ethan, first of all, 200 million, that's a lot.

So a, congratulations, B, these are fake accounts in that they're not real people.

Are the tweets themselves?

Are those actually AI generated?

Ethan, good to see you again.

Likewise, Aiden.

Really, really, it's a pleasure.

It's really good to see you again.

I originally met you because I follow a whole bunch of accounts that you and company host.

They're these like really amazing, you know, X accounts.

I don't know if you have them on other social platforms, but you know, that's how I get my AI news to a large extent is that I follow accounts like yours.

And so I, it was like I have to have Ethan and, or someone from the team, you know, to come on the podcast.

And so, so glad that we were able to do it.

But for everybody who's listening in, maybe you can tell us a little bit about your background and what you guys do.

Yeah, absolutely.

Well, first of all, I really appreciate that kind words.

It's definitely a pleasure to be here.

So we, we, I would say my partner and I, well, I grew up just to start in Westchester, NY.

I went to Washington University in Saint Louis, which is where I met my business partner.

Now we've sort of been like, I don't know, Co conspirators for a while.

We've started and failed a lot of things, but we were always super passionate about content, social media marketing, what have you.

I would say, you know, AI, we sort of became passionate about that at the same time the rest of the world did.

And you know, we kind of, we initially created this directory called AI marketing directory, which was like, it's deliberately named, it's named deliberately straightforward for like, you know, just performs exceptionally well on search engines.

And we, you know, running that for a little while, we realized a very common problem that most people in the AI face to face, which is just distribution.

And, you know, we kind of figured out a way to combine the, those, the, the best of both worlds.

And that led to us creating those Twitter accounts that you're referring to, which by the way, those three accounts do about like 200 million views a year.

And they, they're all, you know, fake AI personas, like they're with a little comical twist, I guess.

But we, we make an effort to like be as authentic as possible, like when we're replying to comments, D Ming people, etcetera.

And like the BIOS are kind of like funny I guess.

So, Ethan, first of all, 200 million, that's a lot.

So A, congratulations, B, these are fake accounts in that they're not real people, but are the tweets themselves are?

Are those actually AI generated or is that you all doing it in the background?

Yeah, good question.

So both in the beginning they were fully written by us, my partner and I and then as it started to grow and the core business which I'll talk about after started to really grow like we figured out different ways to automate them.

So a lot of them actually are written by AI, but I would say like we pretty much always have a real human like comb it over to like kind of craft it in our personal style, which like for us is like informative, but also like relatable.

Like it's like, hey, like this is cool, but like, you know, like I'm just a regular dude, you know, like, like, you know, like I'm learning this for the first time just like you are.

So that's, I would say that's kind of the workflow for him.

It's, it's pretty cool.

So you know, for, for everybody listening in, I didn't know that these were not real people.

I thought it was just like a real account.

And I, I follow it.

The content is amazing.

And so when I message one of these fake accounts and to say, I mean the real accounts, right, it's just, it's not a real person.

And you know, to find out that Ethan and team were were behind this and a lot of it and now learning that a lot of it, the posts are actually AI generated in the background.

You know, it's obviously real content and you're curating real things that are happening in the world of AI.

And you know, obviously there's a method to the madness and you all have put this together.

But the content is amazing and it it's crazy to hear 200 million views is an extraordinary number.

Yeah, well, I really appreciate that.

And and, you know, so like that the Twitter accounts and of itself have kind of blossomed into a growing revenue stream.

I would say that we primarily use them like it's just like a gateway to our core business, which is just like a creative agency for AI and technology brands that is sort of like being creative from that kind of core problem we identified earlier.

So yeah, I mean, I really appreciate that though.

It's it's great to hear that people like what we post.

The other thing which is really interesting about you guys is like, is it the two of you like how big is the agency?

So, so initially it just was the two of us for, I would say for about like a year or so.

And then maybe like four months ago we started putting people on the content team to help with, you know, writing and, and, you know, generating stuff.

So like at this point I'm somewhat removed from like actually directly producing content.

Like we had people coming up with ideas and stuff like that on the agency side.

We it's mostly contracted part time labor.

We don't have too many full time employees.

In fact, we have literally one.

We at this point we've serviced like almost 50 to between 50 and 100 different companies in varying capacities.

You know, coincidentally, I told you about this last time you spoke, we were building our site.

I think it'll actually amazing, hopefully be live next week, which will help a lot of that streamlining it.

But like, yeah, it's super grassroots.

Like not a lot of people.

It's amazing.

And, and so the, the, the reason I want to highlight this is that there's this growing trend thanks to AI of what we call tiny teams, right?

That are able to do extraordinary things and make revenue that was previously, that previously just required a lot of people.

And so you obviously write about AI, you have the AI directory and then you do marketing services and, and help companies.

And a lot of that is powered by AI and AI work flows.

I, I think this is like the new way of building companies.

And I love the way that you all are doing it.

And so excited to have you on today to share some of the work flows that you're using internally.

So what is the first demo that we have lined up?

So the first one is like a basic 2 step workflow on Brock.

Brock three.

I know everyone's talking about Brock four.

I didn't have time to adapt it for that.

I'm sure it'll produce better output, but I just need to be sure.

So this is we're recording on July 11th, I think Brock 4 was 2 days ago.

Yeah, two or three days ago, something like that.

So this one versus the other one that I'll do after, which is a lot simpler.

We actually, we expect like utility out of this.

It's not meant for like brainstorming or you know what, not like we, we're trying to craft actionable output here.

And I also, you know, want to preface this by saying, you know, it's meant to be done with someone who's not necessarily technically inclined, but if you are, there's a lot of ways you could run with something like this in a, in a major way.

So it, the, the point of it is to identify, you know, let's say we're a company.

I modeled it like, like if you're a company like fellow, but it could be any company, obviously, like if you're looking for influencers in a particular space for, to potentially do a marketing or influencer campaign.

And the so that's the first step, identifying them based on my criteria.

And the second step is just crafting sort of like a messaging template to actually get them to agree to terms or provide quotes or what what have you.

So that's the first one on Brock 3 Deeper Search.

Let's see it.

So this is really cool.

So as you're pulling that up, this is almost like the new way of marketing.

Influencer marketing has been around for a while, but I think with more and more channels dying, right?

So traditional SEO dying in its in its original form, these other user generated channels are starting to become a lot more important.

So what Ethan's about to show us is actually really valuable.

And more and more companies including us at Fellow, we are focusing a lot more on this influencer channel.

So yeah, super curious what you're going to show us.

Absolutely.

I completely agree.

So I'm going to.

I also will say as part of the reason this is super valuable for us is not just because of the power of the tools, but also because we've done this by hand before.

And so while you're pulling that up, the other thing to note is that for for those wondering what is Grok?

So Grok is like a ChatGPT or is like a clod, except this is from from Axe, which is Elon's company.

And so this is, this is almost like a division of Axe that is producing this tool called Grok.

And you can, you can get it on the App Store.

You can search for Grok and download it on your phone.

You can obviously access it from your X account and it's, it's really cool.

One of the reasons I like it and Ethan's going to show this as well, is that it actually has access to the, you know, the X feed or formally Twitter feed.

And if you want to know something like super current and what's really going on, this is like the best data source for for this kind of stuff.

Yeah, absolutely.

And I, I will also add that like it's possible to do stuff like this on GPT, but you know, like you said, Grok is fully integrated into the platform.

So it's just like operating at a much more versatile data source.

That's especially if you're looking to do a campaign on Twitter, It's like you, you, you might as well not look any further than that.

So the, the, the prompts for the the first workflow are a little bit involved, but I'll explain why I set them up the way I set them up.

Sorry, it just slowed my so why did I set up like this?

I set it up like this.

I mean, as everyone knows, like in the AI space, you want to give it a lens, right, for act as AB to B marketing strategist.

And that's important for, you know, people mostly associated with topic.

It's not as important for that.

The LLM can pretty easily ascertain a topic.

It's more important for setting the standard, right?

Like I am not, this is not some exploratory exercise.

I'm looking for actionable output that could drive results.

And that's the main reason that's important.

And besides that, you know, it's just like setting the goal, the, the criteria, etcetera.

So I'm gonna copy and paste that in here to by the way.

Oh and by the way, I'm using deeper search which I don't believe you could access this without just having the basic X premium which is like 8 bucks a month so that's worth noting.

Totally worth it, I think it's such a bargain.

Yeah, I completely agree.

And then obviously if you have Twitter X Premium Plus, you can get Croc 4, which is probably incredibly advanced.

I haven't even used it a bunch yet.

So it's going to take like maybe 5 to 10 minutes here.

It's gonna outline the steps and once it gives me the steps, I'll have a few points about like just very briefly about each one.

So assessing user requests, it's pretty much doing exactly what it's saying it's doing, right?

It's, it's matching my intent with the outputs that is required to achieve that, right?

It's, it's, you know, it's crawling multiple live data sources in order to like ascertain what, what that output entails.

Planning the search strategy, same thing, very much like what it's saying it's doing.

It's matching my intent with the output.

But also it's like getting a bit more like involved in terms of like validating what the standard is.

Like I said before, like, you know, we don't just want like inactive or stale accounts that get inflated engagement.

We want, you know, we want real authentic accounts that are associated with thought leadership and exploring data, user data.

So that's just, it's very similar to that, but it's like maybe a bit of a higher echelon of validation, right?

It's like it's referencing a lot more advanced sources like citations, stuff like that to just assess credibility versus like, like I said, you could do this on ChatGPT.

But like this is part of the reason that Grok is so much more powerful for something like this because obviously ChatGPT can pull from external data sources the same way.

But like Brok is has just a greater grip on the, the Twitter pages itself.

So it's analyzing BIOS, pin tweets, what influential people reposted this account, etcetera, etcetera, etcetera.

Honestly, usually when an AI tool takes a long time to do something, it it, it's, it implies that they're really thinking as they're saying.

So it could be a good, it's a good sign.

I actually like, it's one of the most brilliant kind of UI innovations that I think exists in that just by saying thinking and then just by showing you the kind of stuff that it's doing in the background that it's not, you know, being lazy.

It's not taking a long time for the sake of taking a long time.

It's actually doing a lot of work and it's kind of showing us what it's doing live, which is really cool.

I saw April Dunford in there as Dan Martel referenced there.

So investigating a whole bunch of different options as it's as it's going forward.

So this is really cool.

Yeah, I know.

And like you said, you used to do this manually, right?

If you wanted to do this, you would be doing this manually.

It would take you a lot longer, maybe a good day or more to do this kind of work, yeah.

Probably more and, and that's part of the reason that like, you know, I like, I'm able to set up the prompts so quickly because like we have done this a lot manually and and it takes a long time and it's very, very frustrating.

Like, you know, just the whole process and it's a very practical application of AI.

And by the way, I completely agree with your previous point.

Like I think the user friendliness, like the jargon, that's like a huge, huge leap that Brock I think is doing a much better job of.

But you know to.

Take a step back too, right?

What used to happen is, like you said, a used to do it manually, but B, the way you would have done it, if you're a small team, maybe you would have done it yourself.

But if you worked in a bigger marketing team, what would have likely happened is, you know, the manager would have probably gotten somebody on the team and like made it their job of, hey, we want to do influencer marketing.

Go create a big list of all of the people could who could be influencers for our brand.

And, you know, maybe create a table and, and write in each column how many followers they have, what kind of stuff they write about.

And like, go create this whole thing and heck, spend a whole week on it and then bring it back.

And then I can make the decision.

And this is how teams ended up having so many people, right?

But now you don't need a person to go do this.

This is your intern and and this intern is effectively doing this work for you.

100% And to your previous example, like when you're doing it manually, like I would say 75% of the time isn't even like thinking.

It's like you said, praying the tables, getting the, the, the, the metrics, putting them in an organizable way for whoever you're presenting it to.

Like versus now like you could basically spend 95% of the time thinking like, you know, like who's who are the best fits based on this output?

I was gonna say, even if I extrapolate from this, Ethan, the other cool thing is, OK, so now we're using this to get a whole bunch of influencers.

But see, now I wanted to think about, well, what are a whole bunch of newsletters we could pitch that could include content from us, right?

You'd basically do the same thing.

Or if you wanted to say what are a whole bunch of blogs that we could be a guest writer on, you could do the exact same thing.

So anytime you want to find something and do this kind of a research and we were showing it as it relates to influencers, but this literally applies to everything.

100% you could absolutely like set the criteria.

Also, I will say it's worth knowing like, and this is part of the reason we use, I use deeper searches because like it's using all of those sources you mentioned to, you know, sort of basically validate the list that it's procuring.

And like, you know, so if someone's been sighted like, or, or links to open source projects cited in newsletters or research articles, it's using that to say, OK, this person's an original thought leader.

Like they are influential.

You know, this is person that's worth talking to you.

They don't have some inflated metrics like they didn't go viral in 2021 and it didn't, you know, go viral since the validation piece is like the, in my opinion, like one of the more powerful parts of this.

So it's actually doing a lot more than just finding it's, it's actually thinking about whether or not these are the right people.

And if you put criteria like what you mentioned is you can actually put consistency as a criteria that in your prompt as you're creating these things.

Hey everyone, hope you're enjoying the episode.

1 of the common things that we hear is, hey, you all talk about all these different tools, you walk through these demos, sometimes it's hard for me to follow along.

We decided to do recently was take all the things that we talked about, put them in a weekly newsletter.

We literally list all the tools, we link to them and any demo that we walk through, we break it down step by step, put it in this newsletter.

So it's super easy for you to follow along.

And the other nice thing is that if there's someone on your team that you think could benefit, they're working on that subject area.

They could use a little bit of AI injection into their work.

Well, send them the newsletter, send them that particular episode.

We make it super easy to do that.

It's free to sign up.

All you have to do is go to thisnewway.com, enter your e-mail address, and you'll get this weekly e-mail from us.

Hope you enjoy it.

With that said, let's go back to the episode.

Absolutely.

Yeah.

And like you could see like it's almost at 300 sources already in like 7 minutes that it's fully analyzed and it's it's the power of this stuff is, is true.

Like, I don't even think we're scratching the surface, but it's super, super cool.

And, and we use it for this workflow.

We use it all the time.

Like, you know, pretty much every time we, even if we're not actually going to use the output, like you know, if we're just going to kind of maybe tell people some options or whatever it is, like pretty much every time we get a new bigger campaign customer, we do something like this.

Yeah, this is the source count keeps going up.

It's now at 373 sources and and it keeps going.

Yeah, it's incredible.

And I will, I will also say that, yeah, I mean, like you could do this on Deep Search, which, which without the $8 a month, but like the source versatility and, you know, validation piece is like that's the reason why these, you know, more premium or you could say like advanced tools these LMS offer just so powerful for stuff like this.

Yeah.

And also like, you know, if you were a more technical person, like you could totally like, you know, include more advanced data sources, layer them on like get data from LinkedIn or newsletters or what have you.

Or you can create some sort of automated workflow to deliver this to you automatically.

There's so many ways you could run with this if you actually like understand that stuff which I don't understand it too well.

One other question maybe I'll ask you is like what other kinds of things are you using deep research type queries for?

Mostly like for like in terms of like actual driving value, accomplishing work tasks, not too many.

I mean I use it a lot for like brainstorming lots and lots of data sources at once.

So like if I, if it can honestly be reverse engineered, right?

Like if I, if I have like a list of if someone sends me influencers or newsletters or any sort of like distribution channel, right, I could like essentially reverse engineer this prompt and include those channels and have them validate it for me.

Essentially like, you know, so say like all right, is this like instead of like going to all these pages and being like, all right, like this person has inflated engagement, This person uses bots, this person, whatever.

Like I could you know, kind of have grok do that for me on like a really large scale.

That's like the probably the second most common.

But Ethan, how do you do that?

How does Grok know if an influencer is using bots or it's fake engagement?

How does it figure that out?

That's actually one of the, I would say one of the easier parts for it in terms of the validation.

It's really just like they use, they don't like say this you person using bots, they'll say like, and I'll set it up to say like give me a percentage probability, right?

And they'll, they'll, they'll essentially use proportional like engagement ratios.

So they'll say like, OK, this person got 100,000 views.

Like this is a very basic example and like 2 comments and both them are from people they're following.

Like, you know, like that's a very clear sign that this person is like going to some website and paying 20 bucks to get fake views.

And obviously it can become more nuanced than that nowadays.

They're becoming more sophisticated.

But oh, let's see, it's completed.

But that's like, that's a pretty, you know, common example.

So do you put that in the same prompt so that the prompt that you created you will also say, please also verify that most of their engagement comes from real organic engagements versus from bots.

Is that part of the overall prompt?

Yeah, it it is.

I mean, well, for this second type of thing that like I would do, I would reorient this whole prompt around like not just I I don't want them to discover anything like new.

I want them to just validate what I already have.

So I will, you know, kind of reorient that prompt around like that, that subject, that intent.

So yes, the answer is yes, I would, I would literally put the words like tell me the probability if this is bodded engagement, bodded views, etcetera, etcetera.

Very cool.

Yeah.

So as you can see here, this took, let's see how long it took, 10 minutes and 16 seconds, 524 sources.

It gives me obviously a little summary.

It gives me a list of these influencers with, you know, links to their profile with URLs.

It gives me like a really brief sentence about each of them, like what they talk about their, their, their like main sources of credibility.

And then I think it, it'll, you know, it'll sort them by follower account.

But, but it's already been told that like follower account is not really like what we care about here.

We care about real engagement.

And it'll give an analysis their sources and it'll give like this table.

And like, I could literally like, you know, export this table and like send it to someone.

You know, I could maybe change it up a little bit and I'd make it like more to my personal voice.

But like in general, like this, this type of work is the type of stuff that would take like, like you said, an intern or research assistant, like hours, maybe even days.

And like the problem is like, it's not that like allocating days or hours is a problem.

It's like you want to be spending that time thinking about the results versus like just getting the results.

It's just kind of like a waste of time really to like not do do it like in a way like this.

And I mean in my opinion, but I used to do it like with the manual way so I guess.

I mean, this is super impressive.

I mean, you've been scrolling this whole time.

It has an amazing table.

I mean, it did the task all the influencers that you could probably want.

The table is accurate.

You can send this off.

This is incredible.

Like this can replace an entire person.

Yeah, no, I agree.

It's just some powerful stuff.

And, and like, also just like to be clear, like one of the, the subsequent powers and LL NS, like I could just type into this like, all right, like this is amazing.

Like just short it down for me.

Like, you know what I mean?

Like, like if I didn't want to like, but you know, this is like pretty concise as it is.

So yeah, I agree.

So I can go to the second.

The second prompt is a bit shorter.

It's basically just like, you know, taking these results and creating like a messaging template that it will use to like craft a cold outbound strategy.

And I'll basically tell it, you know, this is not to do research.

I want to get quotes from these people to like, get a, let's say a Twitter thread, right?

Like I put your Twitter thread quote for an upcoming campaign.

So once again, like setting the standard here, like this is I need to drive value for my, my employer.

I need actionable output.

And, and that's, you know, pretty important and prompting, probably like the most important part, honestly.

So here we go.

Let's put this in there.

So basically you're saying go through the entire list that you had and go through all of their things that they posted and figure out an angle to outreach to them.

So something that they've said in the past that you can then use to construct a message so that you can, you know, send an outreach to them to basically engage them to help promote your, your company in this case.

Yeah, I specified in the prompt itself like I don't exactly want you to write the message for me.

I want you to give me personalized angles, right?

So like in this case, it's fellow.

So I want you to analyze our product market fit, which you know, and I want you to, you know, take the list you just created.

And I want you to essentially, if I'm a real human, what are some ways I can reach out to these people in a personalized manner to like increase my our response rate, right?

I want them to actually respond like, I don't want to just like copy and paste something and get on with my day.

So like, you know, like that's really the main focus of this, this this part of the workflow, Yeah.

It it, it's kind of crazy because again, you would have done this kind of stuff manually.

And you can see how when I go back to this idea of, oh, it actually replaces the work that a single person would do.

You can see why, right?

It's not just go come up with the list.

It's OK.

What kind of things should we reach out to them with?

And you're you're taking a multi step approach first.

What is the angle that we're going to reach out to them with?

And this stuff is incredibly difficult to do right.

You have to really think about it.

You have to read their stuff and then come up with some creative narratives that you can approach them with.

So it's not like this is like a two second thing.

It it actually requires a bunch of work from whoever's doing the work.

You know, assuming a real human is like actually conducting this workflow.

Not only is it like take a lot of the work work away, but it helps them learn the best ways to reach out to people, the best ways to validate influencers.

That's one of the most exciting parts of AI for me is like, not just like automating work, but it's like helping you think in a higher level way.

You know, about these tasks that you're trying to, you know, do.

And so I think it's completed here.

This one should be a bit shorter output wise.

And so, you know, you gave me some key points, right?

It gave me like an overarching strategy.

They gave me outreach angles, Another, you know, table gave me some angles and like I could obviously play around this and be like, hey, like change the outreach angles to bullet format or, you know, let's not focus on like things like like this person, like, you know, I, I could give a specific focus based on what the angles it generated for me, all sorts of stuff based on what I'm trying to accomplish.

And once again, like you said, like this would have taken so long to do.

Like, like, like the second step was a lot shorter than the first one.

This probably would have taken like.

If it really was a real human doing this, there would have been a lot more thinking time like allocated to this.

We can think about this for so many things, Ethan.

So for example, one of the things for anyone doing PR type work is, OK, let's figure out who's writing about this area, who has written about companies like us before.

So obviously we're doing this for influencers on X, but, but this applies to, to writers in PR, blog posts, newsletters, YouTube creators, like whoever you want to find.

And you can even maybe make the, the prompt even more complex and you, you can put in things like do a conflict check, like are they a fan of one of our competitors?

So maybe that's not the person that we want to engage with.

Or, you know, whatever else your criteria is, it would have taken a lot of time and effort from a, from a human to do this.

But as long as you know what the criteria is, like you said, up level your thinking into the strategic area.

Really think through those right.

Great prompts and instructions and then the AI just does it like magic.

Absolutely.

And like, look, like I think we can both raise this extraordinarily powerful technology.

I will say, though, that like the best way to use it, at least for stuff like this, is to like, look, you know, it does a lot of the thinking for you.

But like to actually think about like, what you're trying to accomplish here and give it as much information as possible.

Like, you know, versus like if you're typing something to a real human being, writing some essay text that's not getting read, but like if you type some essay text to an LLM, they're reading everything, you know, like, so like it's, it's detailed is great, obviously within reason, but like, it just forces you to think about things that maybe you wouldn't have thought about in the past because you're so busy creating a table with follower engagement numbers, you know.

So I agree it's it's incredibly powerful and it saved me days worth of time like in the last couple of months alone.

This is super cool.

Thanks for for showing it to us.

Hey everyone, just a quick pause on today's episode to tell you about my day job.

In addition to this new way, I'm the CEO of a company called Fellow dot AI and Fellow is an AI meeting assistant.

It joins all your meetings.

It summarizes them, tracks the actions and the decision and does that better than any other tool that you've seen.

We've spent a ton of time making sure that the meeting notes and summaries and action items that come out of Fellow are the most accurate, the most precise.

It beats any human you have to try it.

But in addition to that, what makes fellow different is that it is the first AI know take care built from the ground up with security and privacy in mind.

This means that you can use it for all your meetings, not just the customer facing ones, but also the sensitive ones.

Things like one on ones and executive team meetings and those QBRS and everything in between.

It's got really good judgement.

So for example, if you start a meeting and you're talking about some social stuff that you don't really want on the record that's going to get emailed to everyone afterwards, Fellow just knows.

It just doesn't include those things.

Or say you have something that you talked about and later on you realize, oh man, that shouldn't have been there.

Makes it really easy.

You can go back to the meeting, select that part, delete it, and then it's gone from the record.

It connects to all of the other tools you use in the organization, whether that's Slack or Asana or a HubSpot or a Salesforce or Linear or Jira or Confluence, whatever it is, Fellow integrates with all those things.

I think the best part is that Fellow also acts as an AI chief of staff because it sits on all the meetings and the conversations that you have access to.

You can ask it really cool questions such as what are the biggest opportunities in my company?

What are the bottlenecks in engineering?

And Fellow just pieces together information, sees trends, and they can answer those sorts of questions.

He can even do things like, hey, based on all of the one on ones that I've had with this particular person, can you create them a performance review?

And it can do things like that too.

And the Sky's really the limit.

What I really wanted you to do is have the opportunity to try out fellow.

Check it out.

And we're making a special offer available to all our listeners.

So just go to fellow dot AI slash this new way to try fellow.

There's a discount code in there for you if you decide to continue with it.

Either way, I would love for you to try it and let me know what you think.

And with that said, let's go back to the episode.

So, Ethan, what are we seeing next?

So the next one is a a much more basic workflow that I mainly used for like brainstorming, like you can generate output, but I don't as a content creator, I don't typically rely on, you know, automated external tools to produce content.

I help it use it to help me think through, you know, content ideas.

So for this, I'm going to use if in Shachi BT obviously it's many people know they have a GPT store and I'm going to be using the VDGPT to create a very short like 52nd video about like AI news.

And I'm going to also have it like maybe write a script and you, you know, automate the voice behind that script and also, you know, maybe analyze it across popular trending templates in the similar space, you know, to like increase the familiarity with the audience.

So I just want to show up the power, like how quickly you can just create a video and really use this if you do have those skills to create a better video.

Wow.

It's it's like you're basically gonna replace a production studio with AII mean.

I'm excited to see this.

Yeah.

And by the way, so this requires GBT 4 which you can access with GB, you know, track GBT plus which is like $20.00 a month.

So what is Veed?

So Veed is basically a GPT.

It's just it's almost like an extra tool that exists and chat ChatGPT can connect to it and it performs a certain task.

In this case, Veed actually makes videos.

Yeah.

So Veed is a company that we actually work with.

I would say their their most interesting pitch besides generating videos is like they they help create avatars actually that like, you know, well, so if I create a script for this video, they'll have a fake human being that looks super realistic, like saying the words I am doing due to like the lag that I'm experiencing all that stuff lag that I'm experiencing and all that stuff.

I think I just want to show you like even if you click the the no avatar option, like it's still like like very like I guarantee once you see it, you'll be like, Oh, I've seen videos like this on my feed before.

Maybe not about AI news, but like, I don't know about like food reviews or like a review of like like different like water parks, whatever it is like, you know, so veed is like a competitor to like cap cut, but you know, they're I would say they're way more advanced, like, you know, with the avatar stuff, stuff like that.

So.

I'm excited.

So this is a very basic prompt and like as you can see, it's just like create one minute video and 9 by 16 aspect ratio, which is the which is basically mobile phone aspect ratio.

So latest updates in the AI chatpot world figure the ideal viewer is an early stage founder and write a short script reflecting that.

So here we go.

So it's literally just that, and you can always type this in yourself.

It's amazing because a lot of the video degeneration tools that I've tried, they don't do a minute all at once.

So this is really cool that it's going to create a one minute video.

Yeah, so it's exactly.

So it's going to give me a few brief questions.

So I'm going to say hype driven because that as anyone who posts AI content knows, that's what does the best, especially about chat bots, let's say specific, yeah, specific updates and let's say opinion, let's say educational.

So because that's that's going to be the easiest for that to produce.

So let's do that answer those brief questions.

So here's the script.

It's giving me the script.

I can obviously edit this if I wanted to.

And then now I'm just going to click.

So let's just assume that this is perfect for me.

I'm going to click, you know, type continue.

It's it's really cool how this works.

So first you say what you want, it generates the the script.

You can comment on the script if you want and then afterwards you know if everything is good, you just tell it to proceed and it is going to create this video for us.

It's super powerful because like obviously there's tons of AI tools out there and like just having it integrated into like a platform like this is so, so productive.

Like I don't need to go to Ved's website to, to do this.

I mean, it'll eventually redirect me.

Like right now there's a second prompt afterwards, but I was trying to show you what it already generated just from that.

And some of these things, it's interesting, we've also gotten used to instant loading of everything, but these tools are actually doing some crazy things in the background.

So they're actually doing a lot of work.

So it's not loading, it's thinking and it's working.

Once it creates a video, I have a, a bunch of options like basic options, avatar voice only.

I, I mean, I'm gonna click back to the hours to show you like what it looks like.

It takes some time for it to like match up the the audio.

So just for this, I'm gonna we'll play that.

I wonder if it'll go quicker if I do voice only.

I could obviously change the language, the music in the background.

I, I, I think this is, this is cool for people to see, right?

So we are working with kind of cutting edge things.

And so they're not perfect.

You may have to tweak them, you may have to reload the pages, but they do work.

And once you learn how to get them to work again, you're doing the work of an entire production studio.

So I think this is all part of the game.

Yeah.

And it's important, like you said, to have contacts like if you were doing this without any of these tools, like, you know, waiting 5 minutes for it to load is you'd be spending a lot more time than that, so.

What's your perspective, if you had to guess?

So a lot of brands, consumer brands advertise on say Instagram.

From your perspective, how much stuff is now AI generated from the things that we are viewing?

Well, that's a, that's a great question.

I mean, I definitely think behind the scenes, a lot of these platforms are actually, this might sound a little, you know, Y2K like, you know, crazy.

But like, I think they're like potentially trying to in certain niches, push content creators out and just replace it with their own AI content to almost like circumvent sharing the revenue with them.

I think that's definitely something that's happening behind the scenes, but also like just from actual creators utilizing these tools.

I think it's very niche dependent.

I think, you know, if you're just like sharing up news updates, like stuff that like you're just like, it's not much personality in it, you know, not much like really required familiarity.

Like there's no brand, there's no, there's no creator loyalty.

It's just informative.

It's algorithmic.

I would say like an exorbitantly high percentage on platforms like TikTok, Instagram, YouTube shorts.

But once you get into the weeds and like someone's creating a, a personal brand, a you know, a familiar face, you know, it, it, it could vary a lot more.

I guess you guys can't hear that audio, but it's basically reciting the script and this female voice.

Yeah, so this is great.

So what what we're seeing here is that this is the video.

So for whatever reason it we can't get it to play right now, but we are hearing the audio in the background, which is which is kind of cool.

I'm seeing some some captions.

So the the captions are also appearing.

So what we're going to do is like, again, so this is I'm glad that this happened because these tools are new, right?

AI and all the these tools are new.

They do work, but sometimes they can be finicky.

So what we're going to do is we're going to include a link to this video so you all can can watch it.

And then hopefully we're going to obviously include a link to Veed and instructions on how to do this yourself with a custom GPT and, and all the stuff that we showed with Grok as well in the instructions and then newsletter for this new way.com.

So make sure to sign up and we'll send you the link for those things.

Yeah, absolutely.

I guess we won't go through it today, Ethan, but after this video is generated, what would you have done next?

Yeah.

Well, so the second step, what I was going to do very briefly was just say like, all right, like analyze, like here's my niche, here's my target audience.

Like, you know, find me the best performing TikTok or Reels templates that are performing well for like popular creators in the space.

And like the point of this, you know, just familiarity.

It's not just like algorithm performance, which is definitely important, but it's like, you know, when people see something that has that is consistent with a template they've seen from like, let's say a creator like Alex Hormozi or who's obviously not an AI creator, but he's a recognizable face, it just creates credibility.

It creates context.

And it's like it decreases the likelihood they're going to like swipe up, which, you know, is how these algorithms work.

The most valuable metric for, you know, being pushed to an algorithm on a video first short form video platform is watch time.

It's more valuable than engagement by a lot.

And so like, that's why, you know, like high resolution, you know, like those types of things that people a lot of times overlook because of all these tools like is just way more important than people think.

So I use it mainly to like, I don't actually most of the time use these videos and like post them because like we have creators that, you know, create videos and also in our space, what performs best is usually like screen share videos and stuff.

We don't create this type of content, but I use it to like really just brainstorm like my messaging, you know, what types of hooks are performing well or what types of like, what are the different ways I could frame this to like reach the particular target audience I want to have, right?

And like, you know, based on the output, based on the templates it gives me, based on the script, I could see how the AI is thinking about it.

Like it's saying, you know what I'm saying?

So I use it mainly for brainstorming.

I think it makes all the sense in the world.

This has been incredible.

I mean, we've gone through so many things, lots of use cases.

Ethan, I'm sure a lot of people are going to want to get in touch with you.

So if they want to do that, what's the best way to for them to do it?

Well I could provide my e-mail to you or my LinkedIn cuz our site unfortunately is not live yet, which is which is very, you know, lackluster by me.

But you know, if it ends up being live by the time we send out the newsletter, so we'll obviously include it in the newsletter.

What is the social profile you want them to find you on?

Yeah.

So you could just you could, you could either e-mail me at ethan@marketingguys.co or, you know, you could follow this dude likes AI on Twitter or this guy knows AI on Twitter.

And obviously I'm on my I'm on LinkedIn.

My name is Ethan Hochberg.

Well, thank you so much, Aiden.

It's been a pleasure.

Yeah, this is great.

Thanks for doing this.

Yeah, no, I I really enjoyed it so, so thank you.

And that's it for today.

Thank you so much for tuning into this episode of this new way.

If you like the content, be sure to rate, review and subscribe so you can get notified when we post the next episode.

See you next time.