Dan Beresh: [00:00:00] Welcome back to the Fide Podcast. I'm Dan Beresh, founder of Fide, and as usual, I've got Joe, our creative director here, and we're stoked to talk today about AI. Of course, that's the, you know, buzzword of the day on every single podcast, but where we wanna talk about it is, let's talk about creative production, actually creating visuals, videos, assets, et cetera, using the new AI tools.

What do you need to think about when you're using those tools, even as a client? You're hiring an agency and they say, "Hey, we're gonna use some AI." What are the kinds of guardrails that you need to put a- a- alongside, and what kind of expectations can you have? A lot of people coming to us saying, "Hey, can I get this for a quarter of the price?"

I don't always think that's true. So Joe, I'm stoked to talk to you today about something we've been doing a lot of work in, which is this-  

Joe Simone: Mm-hmm ...  

Dan Beresh: AI generation.  

Joe Simone: It feels like the Wild West at the moment. There's so much that you can do, so much more that people think you can do, uh, and then, you know, how to get it all done lies in there as well.

So it's, um... I [00:01:00] don't think it's as cut and dry as people assume it is. There's a lot of, like, variation that you have to account for, um, when working with AI, so yeah, there's, there's lots to talk about.  

Dan Beresh: Let's start right at the top. When you think about an AI project, if you were putting yourself in the shoes of one of our clients, what would you want them to consider to weigh the pros and cons before saying, "Yes, I'm gonna pull the trigger. Let's do this with AI"?  

Joe Simone: I think I would want them to have an understanding about what is possible. I, I know I kind of entered into the AI space thinking like, wow, I'm used to being put inside of this, like, creative constraint box, which is, kind of helps with my creativity. But then when I went into the AI space thinking, oh, well, now I'm kind of floating in space and anything is possible.

And while the underlying truth to that [00:02:00] is like, yes, most things are possible, how things get built and how things come together matter a lot in terms of what you get as an output. And what I mean by that is, you know- You can have this pie in the sky idea and be working towards it, but one iteration in the AI tooling that you're using can throw a huge wrench in the gears, and you can spend a lot of time trying to fix that.

Or you can kind of take an approach of, well, maybe we have to, you know, to quote some writers, kill our darlings here and not obsess over getting this shot perfect, but getting a shot that works and tells the story correctly. I think that's the, the main thing that I try and tell clients is, don't get married to an idea.

Be ready to roll with it, and have a, an end goal that we work towards, but how we get there is gonna be a lot of pivoting, I would say.  

Dan Beresh: That's exactly it. I think you have to have your eyes on the [00:03:00] prize, which is your concept, which is your main goal, rather than y- you know, traditionally, an agency would show you a storyboard, and if they're filming something, they're gonna make the shot as identical to that storyboard as is possible.

With an AI workflow, you look at the storyboard and say, "That looks great," and you get married to it, and say it's a, you know, kid playing soccer and he's so cute, and suddenly when it gets to the AI generation, that kid is in a different-looking stadium or a different-looking field on a different-looking day.

And yes, you can modify the day, but the minute you modify the day, suddenly it modifies what his jersey is or her jersey. So those are the kinds of things that you have to kinda sacrifice. And to me, you know, what I was always taught is you just have to tell a good story.  

Joe Simone: Yeah.  

Dan Beresh: And, and yes, you want it to be high quality, but I think it is a matter of saying, all right, if I'm gonna use this AI tool stuff, I am gonna have to be much more comfortable with, to your point, that creative variation.

And I think where really good [00:04:00] studios and what we're doing now is trying as best we can to create that continuity and, and ensure that it is, you know, on brand for the, um, for the client But what on brand means might be a different look than maybe you had in your mind's eye or in your heart, and that- Yeah

that's something I think you have to accept.  

Joe Simone: Exactly. And I think, you know, when you, when you really do think about traditional film production, it's, it's kinda the same, right? It's just a different way to get there. Like, you could be on set one day with this grandiose idea that you're gonna get this particular shot, and everybody's ready for it, and then the weather changes and you have to pivot.

You just ... It's that same mentality of being able to roll with the punches, and like I said, have that end goal and just keep working towards it, even if you have to pivot how you get there.  

Dan Beresh: There's this implicit expectation, I think, that you have this beautiful magic tool where you can, as you said earlier, [00:05:00] do anything you want.

Sky's the limit. But what you can't do is have that super granular control, and there's an assumption- That's it. Yeah ... I think that we make that, oh, we can make anything we- It's like, yeah, you can, but you kinda have to go by, you know, what the model is gonna give you.  

Joe Simone: Infinity possibility, not infinity control.

It's, it's very loose controls. And there are things you can do. There are steps you can take to get a little bit more control and keep that continuity going, but, you know, it really is like a dice roll. Like, I've sat for hours just trying to generate this one particular shot, and it's just, for whatever reason, feels like the person needs to be grunting while they're doing the dishes or something.

It just inserts some random artifacting, which is ... I mean, it's fun to play around with, but it, it can be tedious and frustrating when you're really trying to work towards a deadline.  

Dan Beresh: You were generating that pirate the other day, and, uh- Yes. The pirate-  

Joe Simone: He's making me laugh ...  

Dan Beresh: the pirate just won't stop grunting, and I just, we just [00:06:00] don't know why.

Like, he can't-  

Joe Simone: Grunting, burping, I don't know. We're- I don't know what AI thinks about pirates, but-  

Dan Beresh: S- uh, yeah, we know now. I think that's the other thing that you mentioned, is the amount of effort and just how much iteration has to go into it, and that's, I think- Mm ... something that is really important to understand.

So if the number one thing to consider when you're thinking about an AI project is that you may not be able to get exactly what you want visually out of it, and that's what you're getting for, you know, what might be a, a, a, a, a cost difference, for example, or a timing difference or effort difference. The second thing is really about understanding that it is a lot of effort to do this.

There is a huge amount that goes in behind the scenes. And so you should not expect that you can get a $500,000 h- you know, commercial for $4,000 and made in, like, a day with AI. And I've, I know I'm, I've harped on this on LinkedIn, and I've said it on the podcast before, like, so many LinkedIn bl- bros are saying, "Oh, look at me.

I made this great AI [00:07:00] video with $0." It's like, yeah, you spent 100 hours on that and generated 10,000 clips.  

Joe Simone: Yeah, time's not free.  

Dan Beresh: And this is the result of your, you know, your masterpiece.  

Joe Simone: Something else that I, I find clients tend to believe is, like, AI does all of it, when once again, AI is just one piece of the puzzle.

You still need editors. You still need motion graphics or VFX a little at a time to marry everything together. You still need to color it correctly. Like, there's still a lot in the background. It's not just like, type in my prompt, I get my video. It, it's more involved than that.  

Dan Beresh: And I think most importantly, and what's gonna become, I think, the number one most important thing, is that you have to have a concept.

You have to differentiate through a way into the story that you're trying to tell, to the message that you're trying to get across. And that, yes, of course, we can use AI to augment, um, but I think what we're gonna start to see is, uh, one of the, the major hurdles I see clients struggling with is [00:08:00] the pressure from on high to deliver tangible marketing outputs, like a video, like a social post, like a campaign.

Mm-hmm. And as that pressure increases year over year, there's less and less room to think about the concept, think about what messaging pillars should be, and think about how to, to go into that. And I think what we're gonna start to see is a lot more AI-generated, "This looks beautiful," and, uh, it's an incredible shot that, you know, would never have been able to be achieved without, you know, the incredible levels of VFX or whatever, but it has no concept.

It has no idea that holds it together, and that idea, I think, is, is the new currency. Um, I think a year from now we're gonna look back and say, "That's what's, that's what's gonna differentiate you." But anyway, we're here to talk about actual AI- Mm-hmm ... production, so let's, let's talk about that. So, so two things to consider.

First thing is to make sure that you're gonna be creatively flexible. The second thing is to understand that there is a huge amount of effort that goes in behind the scenes, that [00:09:00] you need to understand that it is not gonna cost $0. Some, some situations it's gonna cost just as much as a real shoot would've.

Maybe you're gonna get a little extra VFX out of it.  

Joe Simone: Like anything, there's levers. You can pull a bunch of levers, and depending on what levers you pull dictates the output. Yes, you can do things fast and cheap, but is it gonna be great? Maybe not. It's gonna look, probably look like AI. Or you can spend a lot of time and effort into making it not AI, but that kind of balloons cost.

So like, uh, like there's, there's give and take for just like any other kind of production that you're working on. So it's, it's- I don't think you should see it as this is the fast, cheap way to get a video out or content out. It's just a different way to do it that kind of opens up avenues that might not otherwise be possible creatively.

Dan Beresh: One thing I will say that you should anticipate if you're [00:10:00] thinking about a project like this is a different type of workflow. And something that I'm excited about, um, which I think is a different way and is a more exciting way to produce content, and I don't know, this is for me, Joby, maybe this isn't the same way for you.

It- what I get excited about the idea that let's say we're working a comedy spot- Mm-hmm ... and we have 100 ideas. We could make a V1 of every one of those 100 ideas actually come to life in the way it exactly would be in the end result. And so- Yeah ... what's really cool about that is now instead of saying we've gotta make choices in pre-production, we can actually make choices in post-production.

Joe Simone: That you can see.  

Dan Beresh: Yeah, exactly. That you can, like, this is how the, you, you know, I'm gonna call them an actor, for lack of a better word, this is how the actor in the, in the AI clip performed, and this is what we think is gonna be the funniest stuff. We'd done that on a recent project- Mm-hmm ... where we, we took a ton of ideas, whittled them down into, I think, what is it, three or four now?

Joe Simone: [00:11:00] Yeah. It, it's like playing with puzzle pieces and see what sticks, see what, see what works the best. You eliminate the guesswork, right? You can actually generate, you can see- Yep ... you, uh, does this fit? How, how does the pacing work? Okay, if we add a little bit of beat here, then maybe this one, this other bit that we've created can, uh, can slot in.

And, like, you can play with the timing. It just, it makes it a lot more tangible and visual so that you're not kind of guessing, and allows you to just keep iterating and trying and trying and trying, whereas typical production, you just, you know, you only have so much time in the day, you only have so much budget, you gotta pick and choose.

Dan Beresh: Yes. But equally, I think what's really interesting is this infinite ability to iterate also introduces the problem, because I think the other thing that you need when you're supervising any kind of AI production, either as the client or someone even inside an agency, is you need to have decisiveness.

This is good enough. It's, it's where we're gonna land.  

Joe Simone: You can go forever.  

Dan Beresh: And it doesn't necessarily get better.  

Joe Simone: No, [00:12:00] it's just different. Um, yeah. You, you really have to kind of... I, I feel like good enough is maybe the wrong language. Yeah, yeah. That's fair. It's like, does this work? Yeah. Does it look great?

Yes. Sure, we could tinker all day, every day, but, you know, we, we'd never move forward that way. So you kinda have to put some guardrails on yourself to say, "Okay, this works. Let's go. Let's move."  

Dan Beresh: I think the language that I would suggest using is, does this serve that initial idea, like we talked about at the very beginning of this podcast, and is it building toward that in a way that feels professional, that feels polished, that feels on brand, that, you know, if it's funny, and it's, it, you know, it is actually genuinely funny?

And then, then you, you move on and you can say, "Hey, let's, you know, let, let's put this together." I think the other thing, uh, the interesting thing I saw is some research around the fact that consumers don't actually care whether the content was made with AI or not. Really? This is from, um, I think it was The Drum, and yeah, they, they [00:13:00] were sur- they surveyed a number of people and, and the result was when people are told that something is AI or asked to pick out if it's AI, they get really picky and they say, "Yeah, that's totally AI, and it's schlock and it's crap," right?

Joe Simone: Right.  

Dan Beresh: But if they're not asked, if they're not prompted, if they're not told, they'll just watch the content as content. And I think it comes down to, then, the fundamentals, right? Is the story well-told? Mm. Is it interesting?  

Joe Simone: I'm- Just wondering though, like, uh, is, does that mean, like, w- don't, don't tell the audience that this was made with AI so we don't get kind of an adverse reaction that we're, we might be expecting?

'Cause m- the research that you, you're quoting, I haven't read, but just from my time online as an internet participant- ... it's, it's very divisive. It's- It is ... crazy divisive. Yeah. And depending on where you go, you'll have, you know, people that preach the gospel of AI, and then you have people [00:14:00] that want to burn it to the ground.

So, uh, like, what's the line there? Is, is it ... I think you should be forthcoming that, you know, AI was used in this production, but to, like, to what extent do you have to do it? I'm not sure.  

Dan Beresh: I really loved what the Cathay Cargo did in their recent ad that they created with AI. At the very bottom, you know, uh, in a car commercial or something, uh, or in an, y- you might say, uh, closed course, professionals- Mm-hmm

do not attempt, or in a digital more ad, you might see screen images simulated, something like that. I think it's the same. I think it's you- Disclaimer ... you use some language that just says, "Hey, we used AI." And it's, most of the people won't even read it, but it's there if you want. It's being upfront about it.

Uh, my view, a- and, and this is so interesting too because so many people are incensed by AI, and f- my friends are saying, "People are gonna be out of work," and, you know, "Studios like Dan's are gonna be gone in five years," [00:15:00] which by the way I, I think completely untrue. I think we have to move with the times, and I think that, you know, you just have to be upfront, to your point, and say, "Hey, this is used with AI."

And then I think people just relax, and they just wanna be entertained.  

Joe Simone: Yeah. I think, you know, we work in the consulting space where, like, AI is front and center. It's the first thing you read every day because, you know, that's what's new. Consultancy is at the forefront of tech and, like, how to leverage it, how to make it work for you.

And I think everybody outside of it doesn't necessarily have the view or insight that we get into, like, what AI is actually doing. It is that scary thing that's gonna take my job, when realistically, the more you start to work with it, and I've heard a lot of leaders say this, and it does ... You know, it kind of s- sounds repetitive, but it is, it's repetitive 'cause it's true.

It's like AI is not here to replace your job. It's here to help you do your [00:16:00] job, and the sooner you learn how to do it, the sooner you'll benefit from it. The more you put it off, the sooner somebody who does know how to use it is, is gonna take your place.  

Dan Beresh: That's the person who's gonna take your job.  

Joe Simone: Yeah.

Yeah. It's not a, it's not a robot. It's not an AI algorithm. It is somebody who understands how to use it to their advantage. That's the one that's gonna take your job.  

Dan Beresh: I went to Germany last week. I was looking through these beautiful Christmas markets, and I came across this handwork store. Mm-hmm. And I walked in, and everything was beautiful inside.

And I just, I wanted to buy everything in the store, and I picked up a cup, like a single, you know, ceramic mug, and it was 35 euros And it reminded me that when machines come for the work that was previously done by hand, there probably is still a market for handmade work. Mm-hmm. And that market is for a very niche and specific consumer that [00:17:00] wants something, that values something that is done that way.

And I think that that's gonna happen with the work that we do, that most of it is gonna end up becoming a little bit more mass-produced. There will still be, uh, you know, for the very, very best of the best people, that kind of hand work equivalent of- Mm-hmm ... the production that we, that we do. But that if, you know, you're sitting here listening to this podcast as a marketer and you've got a fixed budget, and you have to make an impact, and you have to move the needle, I just cannot see most of you making the choice to buy a $35 mug that's a single mug.

Joe Simone: Yeah.  

Dan Beresh: You need a set of mugs to, to feed your family, you know?  

Joe Simone: Mm-hmm.  

Dan Beresh: That's gonna be the challenge. And so I think, w- you know, you, you, we all, we all kind of have to, have to adapt. And it's very interesting that you say, Joe, you- you're right. We are really at the forefront of the technology, [00:18:00] technology change because of the types of clients that we serve.

I d- I wouldn't say that looking back on it, I would want everything that I buy to be perfectly handmade. I kind of actually do see some advantage in, you know, having, let's just use the mug example, like, four completely matching mugs that are made in an identical way. Like, there are some, you know, advantages to that, but it does, it does change the way we all have to approach both buying and selling these kinds of services.

Anything else that you would recommend in terms of what to think about when you are considering whether you're gonna buy an AI, an AI creative project?  

Joe Simone: I would say you would want to take into account, um- The risks involved. And what I mean by risks is not, you know, not necessarily using AI, but if you are representing a company and that company has a whole [00:19:00] legal department, you have to be prepared to work with that legal department to get this thing done.

'Cause th- like we said at the beginning, the, it is the Wild West at the moment, and the, the legality of everything is unclear. You have to tackle that upfront, because if you wait till your thing is halfway produced and you've invested all this money, I mean, we've seen this ourselves, legal comes in and says, "Absolutely not, not doing it."

You need to be prepared to do your due diligence to see whether or not if it's even the right choice. I mean, I think it is. I think you should be, like, looking into how you can do this stuff, but you need to be proactive about it and not just kinda make it an afterthought of like, "Oh yeah, maybe I should talk to my lawyers about this thing that we're putting out."

Like, do that first before you- ... you not only invest a lot of money, but invest a lot of time.  

Dan Beresh: We handle a lot of risk, uh, departments and feedback from those people, and if there's one piece of advice that I have is treat them like a [00:20:00] partner on your project and not like a guardrail or a stopping block.

The reality is they are a stopping block, but the more that relationship becomes, uh, you know, adversarial, the less I think you're gonna move. And, and you know that you want to do this type of project for so many reasons and, and you're so excited about it, and it's their job to, you know, be reasonable and make sure that you're not gonna put your company in a, in a bad situation.

So obviously none of this is, is legal advice, uh, that we're giving, but what we're s- saying is go get your own legal advice- Yeah ... uh, at the very front of the project and, and, you know, maybe throughout as well. Uh, the, the... Here's, here's the thing with the consideration Especially when it comes to video generation.

All of these v- models have been trained on existing videos. That's how they know how to make videos.  

Joe Simone: Exactly.  

Dan Beresh: But different models have been trained on different videos. You, for example, you've got Veo, and my understanding is that that's been [00:21:00] trained on all of YouTube because Google owns Veo, Google owns YouTube, right?

So that's kind of your simplest path to, like, an easy way to understand what's b- it's been trained on is, is ... And Veo is, is often considered right now one of the safer models. We've got other models like C Dance, uh, and, and others that have been trained on kinda God knows what.  

Joe Simone: Who knows?  

Dan Beresh: Yeah. For lack of a better word.

And there is probably a legal argument to be made that this is a new generation. I mean, that's what the people who are developing these models would tell you, and it's a brand-new piece of art. And, you know, I looked at the Mona Lisa at the Louvre, and now I'm inspired by it, and I wanna make some art based on it.

Do I need to pay those people to, to, to, to make a, you know, my own art? Obviously not. But the, the, the machine, it's not quite that separated, is it?  

Joe Simone: No. It's not, it's not as, you know, as refined or, or as, um ... You know, it, I don't think it's looking at those small details with that much scrutiny. It's [00:22:00] just kind of I mean, like the comparison you made, like I looked at the Mona Lisa and I got inspired.

It's not only looking at the Mona Lisa, it's looking at everything. It's looking at the Louvre and saying, yeah, gobbling it all up and spitting something out. So like, it's, it's tricky. It's, it, it really is tough to navigate, which is why it's important to navigate it cautiously, I would say. Not to completely say, you know, avoid it at all costs.

That's, that's equally as risky. You know? Like you should educate yourself as best you can and get your team involved so that they can help you navigate it.  

Dan Beresh: Yeah, my perspective, Joe, is that we are gonna come out of this kind of nebulous legal area that we're in now, and people will say, no, the things that AI creates are new, net new-  

Joe Simone: Mm-hmm

Dan Beresh: and not copyrighted by, uh, you know, the, the underlying kind of training data, for example. But I'm [00:23:00] sure there are exceptions, and I think those exceptions are gonna be in, you know, your kind of s- more smaller upstart models that might more directly just replicate rather than creating more net new.  

Joe Simone: Mm-hmm.

Dan Beresh: So I think that's, that's the danger. So, so that's the other thing is, you know, pay real close attention to what your agency's using or what your team is using in terms of models. I think you have to have an upfront conversation. Like- Yeah ... what's proprietary for us is our process, not the underlying technology that we use to create it.

I mean, the technology's available to everyone. So everyone- Yeah ... should be having a conversation saying, okay, we're gonna start this process. What models are you using? And let me make sure that those models are acceptable. I, and I think that should be the conversation as well. It's not, hey, legal department, can we use AI?

That, that can't be the conversation. The conversation is- No ... what AI can we use? What's acceptable and what's not? Where's the line? So Joe, let's talk now a little bit about how the sausage is made, how we take the [00:24:00] creative ideas that we have and our clients have and make them real.  

Joe Simone: Mm-hmm. I think it's a lot more- involved than a lot of people think it is.

Like I said before, people think, you know, you just go to your, your video generator, you type in your, your sentence or your couple, like, your paragraph, and you get exactly what you need, or maybe you don't, and then you iterate on it. But it's actually, there's a, a lot more steps involved before you even get to the video part.

Um, at least to be successful, in my opinion. Um, one of the things that we have seen is, you know, you really want to start with a high-quality image. If you can... 'Cause, 'cause an image is not only cheaper to generate, it's a lot easier to customize and tailor when the generation gets it wrong. And that's not to say you even need to keep trying to iterate on that generation.

You can take that into good old classic Photoshop and do what you need to do there, and [00:25:00] then bring that over to a video generator. So I'd say just be mindful of the processes. It's, it's not, like, a one-step thing. It's a multi-step process just to generate one clip, and that, um, without, you know, going into our secret sauce, it's start with an image, customize that image, and then go into kind of the making that image move, turn it into a video from that still.

Dan Beresh: I think it's so funny because it just feels like the same process we've been following for decades, the storyboard, the script, the storyboard, the video. It's, it- Mm-hmm ... everyone's talking like this is so incredibly different and so... And I think to some extent social media promotes people saying things that are like, "This is completely different from everything."

But, you know, in reality is, put that in a storyboard, look at the image, take it, you know, approve it, and then at that point- Mm-hmm ... let's add the motion. Like, this is a traditional animation workflow  

Joe Simone: That's exactly what we did for, for [00:26:00] a video, was start with a storyboard, generate the stills, make sure client was lined on those stills, and then literally pull those stills straight back out of the storyboard and then start iterating with the video generator.

And it worked really well. Um, I think what we had started to try and do was like, like I said, start with that video generator and then, you know, maybe take a still from that generated video, but that's, that's working backwards. It's not a new process. It's just being directed towards different tools.  

Dan Beresh: Yep, absolutely.

And, and putting together the generations, I mean, we're still using very traditional methods of, uh, a, of video editing. I mean, when I say very traditional, I mean last year, uh, you know, la- last year's video editing kind of stuff. And, and- Yeah ... and I think that it, it, it still has to be that because that's what, that's what's gonna kind of put the polish on it that, that is gonna make it feel brand compliant and, and high [00:27:00] quality and all of those things.

So like you've said so many times before, this is a tool that is helpful, that is doing new and interesting things. It's actually not a completely different workflow. It changes the workflow a little bit. It changes the considerations, for sure.  

Joe Simone: Yeah, it's not reinventing the wheel. It's just, it's just slightly different approaches to how you tackle that process.

But the process- Yeah ... largely remains the same.  

Dan Beresh: I think the interesting thing is, though, it becomes a much more technical process, and it-  

Joe Simone: Mm ...  

Dan Beresh: it is bringing, uh, people who were traditionally, you know, very artistic and would be, you know, on a set and thinking through things, uh, with a variety of other people on set, and now it kind of just puts them in a room in front of a computer, and it's, it, it's a matter of, um, iterating in a very technical way.

And with some of the software we're using, you know, some of the API-based stuff is very heavily technical.  

Joe Simone: I needed your help badly on that. Uh, that one platform- Yeah ... [00:28:00] Comfy UI. Comfy UI. I was like, "Dan-  

Dan Beresh: It's, it's  

Joe Simone: tough ... uh, what am I looking at here?"  

Dan Beresh: Well, thankfully, I'm the resident nerd here at Fide, but, uh, you know, I, I think the other thing that's important if you

Again, if coming back to, to being a marketer and thinking about how to work with some of this stuff is I don't think you want that technical person sitting in a dark room in the corner of their basement trying to make your video, even though they're gonna do it for $4,000. 'Cause I think a lot of the other value around what a creative team can bring in any form, internal agency, production company, whatever it is, is that collaboration and the ability to run ideas by people.

And, and- Mm-hmm ... you know, and Joe, you and I will, will send stuff to each other and say, "Hey- Am I o- am I on the right track here?  

Joe Simone: Yeah.  

Dan Beresh: That's the other danger I think that can start to happen, right? Is you're just generating by yourself, you're creating this, this piece of art that's just completely hermetically sealed off from everyone else, and it allows you to go down a route [00:29:00] that doesn't produce a good, you know, a good result because you, you need that, that collaboration and that artistic, uh, this- Uh

sense.  

Joe Simone: It goes back to the, the same theme we keep coming back to. Like, that that's not new. Uh, if I was to work on, like, a big production and then take it away as the editor and just edit away with no feed- outside feedback, it starts to lose all meaning to me, and you just kinda get into this feedback loop.

Um, and it could be good or bad. You could start to s- think to yourself, like, "I've just created a masterpiece," or you could start going down the rabbit hole of like, "Everything sucks, I suck, this sucks," and then you give it to somebody else, they say, "This is great." So it, it's still important to have that collaborative aspect.

You never, you never wanna lean away from that.  

Dan Beresh: I agree, and, and we even saw a project this fall where we were all totally aligned on everything, and, uh, we went down the, the path of, of making this AI video, and we're delivering it to our client who is just absolutely [00:30:00] loving it, and real round after round of feedback, all good, all good.

And then that client shows it to someone else within their organization and they say, "Hey, you know what? This actually really looks like AI, and, like, here are some areas where I think that maybe all of you should, uh, should take a look at it." Mm-hmm. And I, I think that gave us a perspective that was really interesting, because it's very easy to get these, these blinders on.

And so the other thing I'd recommend in an AI project is actually bringing someone else on who hasn't been part of the project, which is kind of a crazy thing, 'cause a year ago I would've said, "Absolutely don't do that." Yeah. They're not briefed, they don't know what the, the, the project is. But get that extra set of eyes because it's, I think  

Joe Simone: it- It's a, it's a pulse check.

Dan Beresh: Yeah.  

Joe Simone: You gotta have them.  

Dan Beresh: 100%. What, what else, Joe? Anything else in terms of how you might Talk a little bit about the way in which we've, we've been working with AI and, and, and creating content.  

Joe Simone: I don't think so. I mean, [00:31:00] I, I, I said at the top, like, it's not just once you, once you've gotten your generation out of the video generator, it's, it's not, okay, it's done.

It's, again, back to the, the, your premieres, back to your Da Vincis. Now you gotta edit it, and now you gotta, now you gotta mix the sound, now you gotta add the sound effects that Veo might not have included, or it did include something like a burp that shouldn't be there. And, you know, it's, it's back to tinkering.

It's just, it's an extra tool in your toolkit that helps with generating creative ideas that you might not have otherwise been able to accomplish, but then it's right back to business as usual, I would say.  

Dan Beresh: Although it is a huge part of the output of the project, if you look at the timeline, like a really solid timeline of something that has a concepting phase, and scripting and all of that, and to your point, editing and all that stuff, the change here [00:32:00] is, like, less than 10%.

Feels like it's a completely different, completely new project. It's not.  

Joe Simone: It's definitely opening up a lot of possibilities, and a lot of doors, and a lot of avenues that were otherwise impossible or not cost-effective. In the year that we have worked with it, it has changed six times over. So the possibilities just keep growing, and it's, it's really exciting to see.

Dan Beresh: Well, I'm excited to see where we get to. I think in six months we might have to re-record this whole episode and delete it, because it'll be just a completely different thing. I do think upfront, the, the, the consideration of just understanding that the creative process is gonna be different, understanding the cost is not gonna be massively less, understanding that you do need to think about legal, that you do need to bring those people in as partners, and that the process is fundamentally the same, but also in some ways fundamentally different, is really important to understand before you, um, you know, you hit the green light and, and pay that agency for that, uh, that fun, new, shiny AI [00:33:00] project.

Thanks very much for listening. Uh, it's been awesome, Joe, as always, to chat with you, and, um, I'm so excited about the new AI projects we've got coming up in the, in the new year. I think we're gonna create- Mm-hmm ... some awesome stuff. And if you're following us on LinkedIn, if you're following me, uh, Dan Beresh, please, uh, take a look.

I think we'll be, uh, we'll be releasing some really exciting stuff in the new year. And of course, if you have an AI project and you're interested in talking about it, don't hesitate to reach out. Check us out at fidecreative.com. That's F-I-D-E creative.com.

‍