AI Ethics: Potential for Harm

Doug Berger: [00:00:00] Welcome to the latest installment of Brand of Brothers. I'm Doug.

Johnny Diggz: And I'm Johnny. Today we're continuing our discussion about the Higgins-Berger Scale of Generative AI Ethics, specifically the potential for harm.

Doug Berger: All right, let's get to it.

​

Johnny Diggz: And we're back. So welcome back, Doug.

Doug Berger: Thank you for having me again- ... I guess. Is that how this works?

Johnny Diggz: Yes. So we've been talking about the Higgins-Burger Scale of Generative AI Ethics, or as we call it, the HBS scale. Mm-hmm. Um, no, HBS. I- HBS. There you go You don't have to scale it. The scale got redundant.

Right. Like AI. Twice. Yes. AI is very redundant sometimes. Um, so we talked about, uh, we gave it a little introduction a couple of episodes ago. We talked about transparency. This is the second, or did we talk about... No, uh, th- this-

Doug Berger: This is, this is the- This is number [00:01:00] two ... this is number two.

Johnny Diggz: Number two. So- So th- th- we, uh, w- this one is, is, uh, the P, which in, in-

Doug Berger: The potential for harm.

Johnny Diggz: Yes. Yeah, yeah. In your, in your acronym.

Doug Berger: Oh, that terrible acronym- Yeah ... that, that's amazing, that everybody should remember about Doug and- It's not, it's not- ... Diggs ...

Johnny Diggz: Doug and Diggs intent to party. Intent to party. Yeah. But it's, it's not in the right order as, as we're doing them. I just realized that. So this is the P, the party.

Yeah. But it's potential for harm. It, it- We're doing it twice.

Doug Berger: Yeah, and I mean, transparency was the T. It was. So- You, yeah ... you know, here we are.

Johnny Diggz: So, um, is it... So this is, this is, there's lots of aspects of what, of these ethical, um, potential ethical pitfalls of using generative AI in business, um, um, especially when you're doing creative work on behalf of clients.

Um, and so we've kind of broken it into these different individual categories- That's right ... um, which include transparency, potential for harm, uh, [00:02:00] displacement, uh, intent, and, uh, data usage, right? Yep. Privacy, yes. Data usage and privacy. So, and some of them overlap a little bit, I would agree. You know, you-

Doug Berger: Oh, I mean, I, I think that they all kind of intertwine.

Sure. Sure. Um, I, I feel like each one is able to live on its own- Mm ... but it also has this symbiotic relationship with the other four of its counterparts.

Johnny Diggz: So the question that, that I have for you is, is it possible for something, uh, like, like generative AI, to, uh, work exactly as designed but still cause harm?

Doug Berger: I would argue that the answer is yes- Okay ... absolutely. That's why it's critically important for there to be some sort of guardrail in place when we're discussing what generative [00:03:00] AI is doing and how it's being employed. So as it relates to each use case, that's really where we need to begin to analyze the, the, the four or five different levels of potential for harm, right?

So starting with basically no risk whatsoever. So what is it that you're creating, but also has no risk as it relates to generative AI?

Johnny Diggz: You know, one of the, um, the most common, uh, complaints I've been seeing, especially on social media these days, is-

Doug Berger: Oh, I know where this is going

Johnny Diggz: um, is i- i- the use of generative AI to create flyers or, you know, uh, stuff notifying people that your band is playing or maybe for a restaurant or-

Doug Berger: What a perfect example-

of, uh, the, the place where there's zero real potential for harm as it [00:04:00] relates to malicious intent, right? You'll notice- ... that I, there have to be caveats put here. Right. But I think it's pretty safe to say that the people that are utilizing generative AI to create these flyers that are primarily put on social media have-

Johnny Diggz: Kind of like digital flyers like that say- Yeah

"We, we've got a promotion this weekend"- Right. It- ... or, "We've got an event."

Doug Berger: Oh, I, I mean, we see them all the time for restaurants, which is a, a, a- Yeah, or birthday parties, yeah ... it, and these are great use cases, these are great case studies really for these microbusinesses, right? That these are the businesses that you're primarily seeing these from.

Are there larger entities that might be using them? Sure. Right. Is there a potential for harm based on that proliferation? Absolutely

Johnny Diggz: I think that we need to do a little asterisk here because we talked about this in a previous episode, but, [00:05:00] um, uh, when we talk about p- potential for harm, one of the subjects that tends to come up is the environmental aspect.

And I think it's important- Mm ... to note that we-

Doug Berger: Grumble, grumble

Johnny Diggz: Grumble, grumble. We, we fully acknowledge that there is an environmental impact on AI, uh, uh, in general, water usage, da- uh, power usage, all, all, uh, land usage, all, all these things.

Doug Berger: Yeah, I mean, we can get into how the utilities are impacted, but at the end of the day, data centers are happening.

Yep. Data centers have been happening. This isn't a new phenomenon. Yeah. Uh, but go ahead. Please continue.

Johnny Diggz: Yeah, no, I just... I, I, uh, also when you, like, I... The, the one, the one, uh, data point that I always think about is when you, when you compare, for example, the water usage of the cattle industry compared to the water usage of, uh, data centers.[00:06:00]

W- you know, u- unless you're a vegan, uh, already, I don't, I, I... You don't have a whole lot of ground to stand on. If you drive a car, if you, if you wash your jeans, you're using a tremendous amount of water.

Doug Berger: Taking a shower. Yep. Yeah. Yeah. If you wanna talk about a superfluous use of, of potable water- There are s- numerous that take place in the bathroom.

Johnny Diggz: The point being that we acknowledge that there is a ethical discussion to be had here. We're just not having that discussion right now.

Doug Berger: Yeah, and I, I don't think it is really relevant for the sake of our scale. The reason I don't feel like it's relevant for the sake of our scale is that our scale is specifically targeting creatives, right?

Right. This is f- to help them be able to use as, as unbiased a measurement as possible without these underlying, these underpinning concepts of [00:07:00] environmental protection.

Johnny Diggz: So back to our AI flyers. So you're a small business, a micro-business as you said. Uh, I mean, we, we know several, you know, uh, small businesses, m- like a Michelin-rated restaurant, local- Numerous, yeah.

Yeah. And, and, um, and they have been using, uh, AI to, uh, generate f- flyer- digital flyers. Um, and, and sometimes you see backlash. You see people putting in the comments- Yeah ... like, "I'm not coming to your restaurant because- Right ... you use an AI flyer." I mean, it's this, it's some kinda, it's a, it's a, it's, it's a form of virtue signal- signaling, right?

Doug Berger: To a, to an extent, but it's, it's valid, right? Sure. So potential for harm is absolutely reputational. Sure. However-

Johnny Diggz: So these businesses are making a choice to, to not hire someone.

Doug Berger: But let's also be clear. But they're also, they're also DIY-type people. Let's also be clear. Yeah. They're, they weren't going to hire someone to begin with.

What ended up happening is that [00:08:00] AI created an opportunity that they didn't have previously, because it costs a lot more per month to have a designer on staff or a marketing director on staff than it does to have a single-seat commercial license for a, uh, a generative AI platform. That said- Is there a risk that comes along with it?

Sure. The bigger issue is the, that 80/20 rule that we continually talk about where unfortunately, because of the sake of expediency, these entities, the restaurants that we're speaking of, the entertainers that we're talking about, they, they need to get the word out. It has to be timely. And the reality is, if you have a marketing department, that requires planning.

And oftentimes when you're talking about these tiny restaurants, these, uh, these independent musicians who busk or, or perform [00:09:00] in restaurants, usually these gigs happen in a very short turnaround time where it's not even really feasible to minimally hire someone on Fiverr or Freelancer or one of those outfits, or finding a, a resource on Threads, right?

It, it, it's just not viable from a time constraint perspective, but it's also financially not viable. How much is a freelancer going to charge for one of these social media flyers? On the low end, on the low, low end, you're looking at 25 to $50. Okay, great. Sounds awesome. How long does it take to get them to generate original layout design?

And here I am, a designer- Correct ... a brand strategist- Right ... talking about utilizing AI in an appropriate environment.

Johnny Diggz: This is your bread and butter.

Doug Berger: It, it is. Absolutely. But you know what? These individuals, they [00:10:00] weren't going to hire me, or another designer for that matter, because resources are slim. Sure.

A, a musician might get paid 150, $300 for one gig. Yep. How much of that percentage- ... are they going to give? They're already spending money on, on the gear that is- Right ... depreciating in value for usage. Right. Um, they're already spending money on transportation. They're already spending money on the actual mode of transportation.

Johnny Diggz: Not to mention the years of, of, of training and experience. Well, it, and then- ...

Doug Berger: and then musicians have their bills to pay. Sure. So is an extra $25 out of that $300 a realistic expense? And I, I can't in good conscience suggest that that should be done for every single one of a musician's shows. It- That's absurd

Johnny Diggz: and their talents aren't in graphic design. Their talents are in, you know, pres- [00:11:00] presumably music. Yeah. And, uh, and so, you know, I, as a musician myself, I find my, if I find it's, it, if I- If I interrupted, and I wanna let people know-

Doug Berger: Oh, I don't forgive you for using it.

Johnny Diggz: I do work with a bunch of designers. Um, but even, even in that use case where I do have access to professional designers, um, would I rather, uh, get the flyer out and get that checked off my list and, and you know...

And here's another thing is, like, I know I could just type out that I'm playing a show- Yeah ... in, in a text, uh, pro- in a text, uh, type social media post. That post can't be posted on Instagram, first of all, and I know algorithmically I will be penalized on Facebook if it's just a text post. It d- Facebook's algorithm just doesn't show it to as many people.

So already I have to add a picture. Historically, I have added just a picture of me, or [00:12:00] may- hopefully there's a picture of me at that venue maybe, but m- maybe there's other people in it that don't necessarily... Like, there's-

Doug Berger: But at the end of the day, there's no absolute means by which to completely eliminate the potential for harm, right?

Sure. It, it's, the only way to completely eliminate the potential for harm for using AI is to not use AI. So- Sure ... it, it-

Johnny Diggz: But can you these days not use AI?

Doug Berger: Well, I, I, I believe- ... I, I think it's pretty safe to say that we've been using AI longer than we've been aware that we've been using AI. It's just the way that the term has been packaged.

I, I wanna focus on, on generative AI here, where the, the, the opposite end Of- of potential for harm is where we see deepfakes. And these deepfakes have- have a- a- a duality to them, right? So there's the side where it's a celebrity impersonation, where they've licensed their voice or they've licensed their [00:13:00] likeness- Sure

which has- has gotten really interesting with the voices of, uh, of James- James Earl Jones, yeah ... yeah, James Earl Jones. Yeah. And then I think you mentioned, uh, in a conversation we had previously, not in the, in the pod, but, uh, offline about, uh, how they're, they're... Uh, like Tom Hanks, I guess, is licensing- He was, yeah

his voice.

Johnny Diggz: Well, he, not licensing it, but he was postulating that if, you know, the- the existence of Toy Stories beyond his life- Sure ... um, kinda scared him, you know? Be- uh, as, you know, i- Interesting. Yeah. He- he, uh, he expressed concern about it. But the reality is that in, you know, in 20 years, we may not, God forbid we don't have Tom Hanks anymore, um- And

Doug Berger: he's a treasure.

Johnny Diggz: He's a, he's a fucking treasure. Um, but, uh, uh, he could theoretically license his voice and, and, um-

Doug Berger: We know Nick Offerman is already doing it. Yep. The- the amount of times you hear his voice in a commercial, the likelihood of him [00:14:00] having actually sat in that studio, it- Yeah ... we know there are platforms where you can at...

I can go right now, and I can, using, I'm not gonna mention the brand, um, there's a platform where we can go and license a celebrity's voice. Um, obviously, there are- are limitations. They p- have put guardrails in place that are- are already contractually protected. Mm-hmm. But then there's the other side.

Then there's the other side where people are not using the licensed personalities. Sure.

Johnny Diggz: And the, or, and this is really potentially harmful in, in, you know, in p- political figures. Um, probably the most, it, you, you know, it w- where we're seeing, you know, a deepfake of, uh, of a presidential candidate saying something, or a deepfake of a, uh, an, a- a- a, uh, political commentator saying something.

Uh, something that [00:15:00] people trust, uh, uh, is suddenly can be perverted into an entirely different message than that person would actually, uh, normally say.

Doug Berger: Right. That misinformation component is one of the biggest potentials for harm. Potentials for harm.

Johnny Diggz: Along those same lines, there's- In generative AI, you can also have, um, you know, if you're generating, um, content, uh, text content or, um, blog posts, uh, manuals, instructions, legal briefs- Right

um, medical, uh, medical, uh, suggestions, um-

Doug Berger: So, so on our scale, we have four different levels when it comes to potentials for harm, and, and you're getting into that moderate risk gray area, right? So that deep fake where it's, it's talking about, it's, it's a celebrity or a, uh, [00:16:00] political figure, um, saying something that they didn't actually say, obviously there is malicious intent associated with that.

Johnny Diggz: Sure. But then-

Intent is a whole different s- section on the scale, but yes. Yeah ... right. But obviously, you know- Potential for harm is there ...

Doug Berger: it's symbiotic. Yeah. So when it comes to where you are now, now we're in that moderate risk component, so where it, th- it's really easy to transcend from low risk with guardrails, right?

So a low risk with guardrails might be a AI automatically generates a, uh, a, a series of articles based on a prompt. But then there's human involvement that creates a guardrail to make sure that the information is factually accurate, it's unbiased, right? But then if you let it loose, and you're talking about things that can potentially harm not just the reader, but also potentially society as a whole, we're getting into this [00:17:00] murky gray area where we would say that that is a, a moderate risk, um, with the potential to mislead.

Johnny Diggz: The, uh, that, that sort of, uh, uh... Area where the type of content that you're generating. So, uh, if you're an individual company and you're using generative AI to post blog posts, uh, to generate blog posts on topics that are related to your business, um, and you post... And, and maybe somebody reviews those posts, and you post that on your website.

Maybe they ed- do some light editing. Um, but gen-

Doug Berger: By at least 20%.

Johnny Diggz: By, by at least 20%. Um, you would... That, in our, in our scale, that would generally fall in the green area, right?

Doug Berger: It would. It- It, it absolutely would. But then when you get into certain topics of conversation, like medical- Oh, yeah ... and [00:18:00] legal, right?

Where it... We're talking about industries where there's licensure that- Right ... is, is required. Um, and, and, and it might even go as far as, like, in electrical work, in plumbing, right? Where you're, you are legally required to be a licensed professional. You shouldn't be allowing a A, a computer that just basically takes pattern recognition and fuses it together to give across a message to inaccurately tell someone how to do, uh, plumbing work or, or worse, medical intervention.

Johnny Diggz: We're seeing this more and more on YouTube where, where scripts are being generated, then the, uh, the images are being generated, and then the actual voiceover is being generated. So you've got an instructional how-to video that has not d- been properly vetted by any human, possibly, and telling [00:19:00] you, um, how to prepare a meal, how to clean a wound, how to prepare for a legal, uh, proceeding.

This is... These are all, all very, very potentially harmful- Oh, the- ... ideas.

Doug Berger: The potential is, is high- Yeah ... and the slippery slope is, is significant. So, um, you know, there, there are numerous avenues by which we look at what, uh, this potential for harm is, but there is a, a qualifier, and that, that qualifier is significant demonstrable benefit, right?

So if we're creating articles that are informative, and maybe it's because we've trained a language model on our own repository that we've already written content for, and then we take it one step further and give it more training data, it's fine, generally speaking, to create [00:20:00] content based on content that you already have created because it's already been vetted.

Johnny Diggz: Can you give me an example?

Doug Berger: No. Um, so it, one instance that I can kind of speak to is, uh, we've used our own internal training data to help us generate creative briefs So what we have done is we have created a, a set of parameters that help it not only reference existing creative briefs in our repository, but it also is analyzing a few documents that help to identify the core components that would go into the creative brief.

So, uh, is it a 100% element? No, because we then go in and make sure-

Johnny Diggz: You don't just turn that over to the client.

Doug Berger: We- no. Absolutely not. Um, because, listen, it- time and again, it's wrong. Right? Mm-hmm. Uh, ChatGPT is notoriously incorrect, [00:21:00] which is why it is absolutely critically important for there to be a human in the loop.

Um, and so that potential for harm is another key, key component. I know we talk about displacement as one of our five metrics in, uh, when it comes to, to ethics, but that is a, a key component, is eliminating a, a person, uh, because it's creating societal harm.

Johnny Diggz: So in speaking of the societal harm, I mean, it- w- there's an argument that says, well, if we, uh, if we continue down this path, um, w- people, the, the, the actual creatives, the creators of the world, will be less incented to create because the, the AI, um, will suck up their creative endeavors.

Yeah. They won't be compensated, and why, why would they want to put their, put their stuff out there if it's gonna get stolen by this, by these data hogs?

Doug Berger: Right. So let's look at [00:22:00] history. Um, I, as someone who, uh, has a background in art history, I'm gonna go to where we've been in the world of art. So over the centuries, we have seen artwork, artwork styles evolve, and at their moments of evolution, um, we have seen controversy.

Uh, I, I especially like to look at what happened during the 20th century. During the 1900s, um, we... Actually, we can go even further back, uh, to, to photography and, and video, um, a, or the early stages of it, and then we end up in the early 1900s with these amazing photographers, and, like Ansel Adams. And then, you know, we, we fast-forward and we have, uh, a, a, people like, uh, I think her name is Diane Arbus.

The, these amazing photographers. They took something that was a tool, and they created an art [00:23:00] form from it. Uh, a- another, another paradigm happened with digital art. So this is in my lifetime, where I witnessed artists begin to leverage digital media in order to create art. When you look at the varying media- The oil painting is relatively new, right?

It's something from, like, the, the 14th, the 1400s. Um, and, and we were mixing pigment. Now we have acrylic painting. Acrylic paintings happened in the 20th century. We did not have the technology. And so people could argue that acrylic painting isn't painting because the, because of the craft. But the craft doesn't make the art.

The artist makes the art. And, you know, we still have these movements today that are from a long time ago. We still have people chiseling stone. We still have people casting bronze. [00:24:00] We still have people painting realistic, almost Renaissance style paintings. So I don't think that the art movements have inhibited artists.

I think what it's done is given them more tools, more media in which to explore, and I, I think it's probably just a matter of time before you see these amazing two-dimensional artists who have explored into, uh, video, uh, and immersive, like Derek Adams. I, I would not be surprised if his next level is to take his aesthetic and somehow infuse generative AI a- almost as a dynamic, immersive experience.

Johnny Diggz: You can actually tell the, you know, generative AI in the hands of an artist versus in the hands of, of, of a, someone who is just typing a prompt, and they don't know what they're, want. The, and the, and the, and, you know, ChatGPT is, is great at taking any prompt. And you say, "I will need a birthday [00:25:00] card." F- and you, you type the, the- Yeah

the date and the time and the theme, and it will generate, uh, a very, very complex, um, sometimes overly complex- Mm-hmm ... birthday image, um, with a lot of aspe- a lot of details that you probably didn't ask for. Sure. Um, but in the hands of an designer, you, he, they can ask for something very specific and get generally what they want, right?

Doug Berger: Yeah. I think the 80/20 rule always applies, though. Uh, so when you... First off, I, when you prompt something, I believe that the individual that's making the prompt is only gonna get something that's about 80% as good as they are, right? Mm-hmm. So if you are an amazing designer, there is a likelihood that you can create something that is great.

Is it gonna be amazing? No. That's the next 20%. That next 20% [00:26:00] is where that designer goes, "Okay, so, uh, AI is terrible at margins," as Simon will tell you. Um, a- and so, uh, not only does it not give elements breathing room, it loves to fill spaces. And so you're- More tokens used, so Oh, sure. What- whatever it takes, right?

To fill those pixels and tokens. Um, a- and so I, I, I think it's safe to say that a great designer is only gonna create good design, and a good designer is gonna create mediocre design, and you can see where this is trending, that it, it... The same is gonna be true with, with a copywriter. If you're not good at writing, your ability to discern whether or not the content that you're getting back from Claude or, or Perplexity or ChatGPT, it- y- you're only going to be as good as what you know, right?

So if you are expert level, you're gonna have more [00:27:00] discerning tastes. That's just- Right ... uh, that's just how it goes.

Johnny Diggz: And sometimes 80%, uh, depending on the use case, is gonna be close enough for what you want. But it, you know, if, if you're, if you're shooting for something that is going to be, uh, delivered to a client on behalf of you, that's...

80% is not gonna be- Yeah ...

Doug Berger: yeah. A- and there, there is one last point that, that I would like to talk to as it relates to potential for harm, um, which also makes for a great segue into privacy and data usage for our next- ... uh, for our next conversation. Um, and that is ownership. So when it comes to intellectual property, when it comes to copyright- Generative AI presents a conundrum, and that conundrum is whether or not you are...

That potential for harm, whether or not your customer can trademark what you provide them, whether or [00:28:00] not it is ownable because of where that, that source material is coming from.

Johnny Diggz: And, and, and where the laws currently stand, if there is, uh, I, I, you know, it's, it's a pretty gray area. Yeah, it's still murky.

It's still very murky, but, you know, m- it's, the law seems to be leaning towards, you know, me- uh, they use the terms like meaningful human, uh, input or, uh, uh, you know, a, a, a certain- That's right ... percentage of... You know, it's, it's very subjective at this point.

Doug Berger: It, it is almost along the same lines as parody, right?

Sure. So where parody, that you're creating a song, for example, and you're changing the lyrics, you have to change the lyrics so much- Right ... um, in order for it to be parody. And I don't know what it is about the music notes, if there are requirements in there as well, but I would imagine there's some degree of, of modification that has to take place in order for it to be, uh, to be considered, [00:29:00] uh, someone else's intellectual property.

So at that stage, again, I, I look to that 80/20 rule. Yeah. So, uh, is it okay for a designer to ideate using AI? Sure. Why not? Is it okay for a designer to then have that AI-generated design and bring it into, for example, Illustrator and automatically turn it into vector? Is that okay? No. Because there hasn't been any meaningful involvement by the designer outside of the use of the tool.

Johnny Diggz: You know, I recently saw a, uh, an image that's been floating around, uh, like a- it, it's kind of a network of, uh, all the connected lawsuits going on between, you know, publishers, photographers, uh, uh, creatives of all types, musicians, um, you know, a- any, uh, record [00:30:00] companies suing the, the big AI. Every, every one of them.

Oh, yeah. They're all getting sued. Oh, yeah. And, um, and in almost every direction, and, um, and that, those lawsuits are not settled. Um, they are, they are still ongoing. They probably will be ongoing for years. But, um, if, if any of those lawsuits end up, and the law ends up favoring the creators over the AI companies, it could open up potential risk and potential harm for any company that has, um, used generative AI- Yeah

for their products, for their, for their marketing, for any, any of their, you know, that, there's a, there's a huge potential legal liability there. So it definitely needs, I mean, I'm, I'm sure all of the attorneys of the world are aware of this, or at least some of them. Um, but, you know, when you're making decision, creative [00:31:00] decisions, this can have long-term ramifications- Absolutely

for anybody in, you know, from the creative agency side down to, you know, in, in the corporate side. Um, but fortunately for- Uh, you know, those, those micro businesses that we were talking about at the beginning, not so much potential downside.

Doug Berger: No, not so much. I, I think that they're generally safe. Yeah. Um, I, I don't think that anyone's gonna come after them for, for intellectual property infringement.

It, it's not a good look.

Johnny Diggz: Well, the, you know, the, the reputational damage of people saying, you know, you'll see people, "I'm not gonna go to your show. I'm not gonna go this, to, uh, you know, this restaurant." I, I think that's s- ma- probably some temporary posturing. Um- I think so ... these are also m- probably the loudest voices that wouldn't go to your restaurant anyway, or aren't going to go to your show anyway.

Yeah. Um, a- I-

Doug Berger: and the same, the flip side is true. These are the same people who wouldn't have been hiring someone to do the work anyway. Right. These are people that are, are on a shoestring [00:32:00] budget and are likely bootstrapping their own operation. Absolutely.

Johnny Diggz: So we, we mentioned that, um, next time we're gonna talk about data usage.

Yep, privacy and data usage. And that's gonna be a big one because this is, we were talking about the, the training data, the, uh, and this also ha- you know, it has lots of potential for harm. It also touches on displacement of, of other people's work and, and some legal ramifications, so that's gonna be a very exciting episode.

I hope you guys, uh, like and subscribe, and, uh, let us know how you, uh, how you guys are dealing with things, uh, ethically dubious in the, in the, the generative AI world in the comments. And, uh, we hope we'll see you next time.

 

Johnny Diggz: Thank you for tuning in to Brand of Brothers. Big thank you to our presenting sponsor, Remixed, the branding agency, along with production assistance from Johnny Diggz, Simon Jacobsohn, and me, Doug Berger. We can't forget music by PRO. Speaking of not forgetting, remember to do that like and subscribe thing and find us at BrandShowLive.

[00:33:00] com and follow us on the socials at BrandShowLive.