511. Staying Human in the Age of AI: Insights from SXSW 2025

Summary

Podcast Summary: The Brainy Business | Understanding the Psychology of Why People Buy | Behavioral Economics

Episode 511: Staying Human in the Age of AI: Insights from SXSW 2025 Release Date: July 3, 2025
Host: Melina Palmer

---

Introduction

In Episode 511 of The Brainy Business Podcast, host Melina Palmer delves into a compelling discussion about maintaining the human touch in an increasingly AI-driven world. Drawing from her experience moderating a panel at South by Southwest (SXSW) 2025, Melina brings together experts to explore the intersection of behavioral economics, marketing, and artificial intelligence (AI).

Panelists Introduction

Melina Palmer introduces the distinguished panel members:

• Oussama Fayed: Executive Director of the Institute for Experiential AI at Northeastern University, founder of Open Insights, and former Chief Data Officer at Barclays and Yahoo. [00:35]

• Joanna Lepore: Global Foresight Strategist with expertise in transformation growth for major brands like McDonald's and Mars Wrigley. Founder of Foresight Insights Inside Group and host of the Looking Outside podcast. [02:35]

• Ben Gutman: Marketing and communications educator, founder of the marketing agency Built and Sold, and author of Simply Put, Why Clear Messages Win, and How to Design Them. [02:35]

The Role of Human Communication in the Age of AI

The central theme revolves around enhancing human-to-human communication amidst AI advancements. Oussama Fayed emphasizes the importance of maintaining the human element in understanding customer behavior and decision-making. [02:35]

Ben Gutman highlights the risk of over-relying on AI tools in marketing, which may undermine the fundamental goal of connecting with audiences on a human level. [06:00]
“When we do too much of that, we're sacrificing our ability to do what is actually the real goal of things... to make that connection between us, between our audience, to fulfill their wants and needs.”

Joanna Lepore adds that AI often merely automates existing processes without genuine understanding, referring to AI as “stochastic parrots” [07:01] and stresses the necessity of keeping humans in the loop to guide AI effectively. [07:01]

Risks and Challenges of AI in Marketing and Branding

Quality of AI-Generated Content

Ben Gutman expresses concerns about the disconnect between marketers using AI to generate content and consumers’ preference for human-authored material. He cites personal experiences where AI-generated content, such as blog posts, fails to engage audiences effectively. [11:02]
“We're putting this kind of unneeded and unnatural kind of filter in between... if you're really trying to build relationships with our customers.” [11:02]

Joanna Lepore reinforces that AI-generated communication often lacks the depth and authenticity of human interaction, leading to decreased consumer trust. [08:42]
“I think AI is making us much worse, not better. We have become more effective at gathering information... but it detracts because you're talking to a human.” [15:33]

Brand Risks and Public Perception

The panel discusses notable instances where brands faced backlash for their AI endeavors. Ben Gutman references Skechers' AI-generated billboard and Pentagram's AI illustrations for the Biden administration, highlighting public reception and ethical concerns. [20:37]
“Everybody looks at that and goes, what am I looking at here? And it doesn't even have shoes that Skechers would sell as part of it.” [20:37]

Addressing AI Biases and Responsible AI Use

Joanna Lepore emphasizes the critical need for responsible AI practices to mitigate inherent biases in data. She discusses systematic approaches to detecting and addressing biases, ensuring AI tools serve diverse populations effectively. [23:00]
“Bias comes by definition. Any data set is a snapshot of the world which will have biases in it...” [23:00]

Ben Gutman echoes the importance of ethics in AI deployment, stressing that biases often stem from human imperfections in data selection and curation. [24:11]

Starting Your AI Journey: Best Practices for Organizations

Joanna Lepore outlines three key strategies for organizations embarking on their AI journey:

1. Data Collection: Ensuring comprehensive and detailed data necessary for effective AI training.
2. Human Intervention: Incorporating continuous human oversight and corrections to refine AI outputs.
3. Decision-Making Boundaries: Avoiding the delegation of critical decisions to AI, given its current limitations. [34:48]

Ben Gutman recommends a balanced approach combining controlled case studies with allowing employee-driven AI experimentation to harness its full potential without stifling innovation. [36:22]

The Future of AI and Maintaining Human Elements

Joanna Lepore expresses concerns about AI's role in spreading misinformation and creating deepfakes, which could erode trust in digital content. [54:40]
“The ability to leverage the technology to create... you cannot believe anything you read, anything you hear, anything you see.” [54:40]

Ben Gutman counters with optimism, suggesting that AI will continue to evolve and integrate into various sectors, emphasizing the enduring value of human creativity and critical thinking. [43:00]
“You make more weird stuff. I like that.” [55:36]

Conclusion and Final Thoughts

As the panel winds down, key takeaways include:

• Melina Palmer underscores the importance of balancing AI integration with maintaining human-centric approaches to ensure successful business outcomes. [56:13]
  “If you consider the humans, you're more likely to have success in your AI endeavors and behavioral economics can help you do just that.” [56:13]

• Ben Gutman advises against replacing human talent with AI, advocating instead for using AI as an accelerator of human capabilities. [41:37]
  “Don't fire the humans. AI is an accelerant on your own human capital.” [40:00]

• Joanna Lepore stresses the inevitability of AI in the knowledge economy and the necessity for humans to adapt by enhancing their unique cognitive and creative skills. [50:50]
  “Teach people how to think in this new age and how to be robot proof.” [50:50]

Melina Palmer concludes by inviting listeners to engage with her for further discussions and to apply the insights from the panel to keep their businesses both AI-savvy and human-friendly. [56:13]

“If you do not pay attention to that human element, people will likely be the downfall of your AI plan both on the customer side and the employee side.” [56:13]

Notable Quotes

• Ben Gutman [06:00]: “When we do too much of that, we're sacrificing our ability to do what is actually the real goal of things... to make that connection between us, between our audience, to fulfill their wants and needs.”

• Joanna Lepore [07:01]: “They don’t even understand the words they’re saying. They don’t even say words. They say tokens.”

• Ben Gutman [11:02]: “We're putting this kind of unneeded and unnatural kind of filter in between.”

• Joanna Lepore [15:33]: “I think AI is making us much worse, not better.”

• Ben Gutman [20:37]: “Everybody looks at that and goes, what am I looking at here?”

• Joanna Lepore [34:48]: “Any data set is a snapshot of the world which will have biases in it.”

• Ben Gutman [41:37]: “Don't fire the humans.”

• Joanna Lepore [50:50]: “Teach people how to think in this new age and how to be robot proof.”

• Melina Palmer [56:13]: “If you consider the humans, you're more likely to have success in your AI endeavors and behavioral economics can help you do just that.”

---

This episode offers a nuanced exploration of AI's impact on business, emphasizing the indispensable role of human ingenuity and ethical considerations in harnessing AI's potential. Whether you're a small business owner or part of a large corporation, the insights shared by Melina Palmer and her esteemed panel provide valuable guidance on navigating the AI landscape while keeping the human essence intact.

Transcript

[Melina Palmer]

Welcome to episode 511 of the Brainy Business Understanding the Psychology of why People Buy. In today's episode, I'm excited to share my panel from south by southwest 2025. Ready? Let's get started.

You are listening to the Brainy Business Podcast where we dig into the psychology of why people buy and help you incorporate behavioral economics into your business, making it more brain friend. Now here's your host, Melina Palmer.

[Oussama Fayed]

Hello. Hello everyone.

[Melina Palmer]

My name is Melina Palmer and I want to welcome you to the Brainy Business Podcast. In March of this year I had the absolute honor of speaking at south by Southwest for the second time. If you aren't familiar with the event, it is a 2 week long 10 days or so festival that takes place in Austin every year and they now have events in London and Sydney. Sydney. In addition to the conference events they have a film festival, music festival, comedy festival, food festival, and more. It is massive. Around 300,000 people attend each year. Yes, I said 300,000 attendees. There is so much content. The session you are about to hear was the panel I moderated called Enough with the Delving and Staying Human in the age of AI and it had about 60 concurrent sessions running at our same time slot just on the official south by Southwest events. There are other events going on all across the city throughout the days of this event from Inc. Fast Company, the Female Quotient and so many others. It really is amazing. Highly recommend it. Fun fact, Kevin Bacon was speaking at the same time as us and yes, many people still chose to come to our session, which is pretty awesome. I am so delighted that south by Southwest gave me permission this year to repost this session for you here on the Brainy Business. It's going to be super awesome. Really quickly before we jump into it, I want to be sure you know that there are links in the show, notes for my top related past episodes and books, ways to get in touch with me and more. It's all within the app you're listening to and@the brainybusiness.com 511 now let's jump right into Enough with the Delving at South by Southwest.

[Oussama Fayed]

The central real theme of the conversation that we are having here today is in being able to ask in this age of AI, how do we communicate better as humans to humans? For me, I am an applied behavioral economist and so working in the space of human behavior, human decision making, understanding why customers buy, why employees buy in, that human element is always very central to the work that I do at the Brainy Business through my podcast Things like that, and I'm going to go ahead and introduce the rest of the panel and the amazing things that they do. First we have on the end is Oussama Fayed, who is the inaugural Executive Director of the Institute for Experiential AI at Northeastern University. He's a chairman of Open Insights, a company he founded in 2008 to activate data for AI and evaluate the data assets along with new revenue streams from data and AI leveraging open source and latest data technologies. Previous roles include Global Chief Data Officer at Barclays Bank, Chief Data Officer at Yahoo. He started his career at NASA's Jet Propulsion Lab where he founded the Machine Learning Systems Group. Who's been to a a session with an astronaut already a few. There are quite a few of them on the list. So he founded that machine learning systems groups delivered award winning advances in AI and ML space programs. He had his time at Microsoft, including forming the Data Mining and Exploration Group and building database components in SQL Server 2000. He's published over 100 technical articles on data science, AI and machine learning and other topics and holds more than 20 patents. Next to him we have Joanna Lepore. Joe is a global Foresight Strategist with 15 years of experience building transformation growth plans inside of big brands. After 10 years in marketing, she moved to build and expand the foresight practice for McDonald's and Mars Wrigley from within their global headquarters. Last year, Joe founded Foresight Insights Inside Group, a network for foresight practitioners inside the world's biggest organizations. And on the side she hosts the Looking Outside podcast, exploring fresh takes on familiar topics. And then right next to me, here is Ben Gutman. He is a marketing and communications educator and executive. In a previous life he founded, Built and Sold, a marketing agency and today he teaches at Baruch College and is the award winning author of Simply Put, why Clear Messages Win and How to Design Them, which is available in the bookstore. And he's doing a book signing after this. He's worked with clients including the NFL, I Love New York, Visit California, Comcast, NBC Universal, and the Nature Conservancy, and his writing or work has been featured in the New York Times, Wall Street Journal, Adweek, and other publications. And in July he's going to be the Executive Director of the Queen's Economic Development Corporation. And now you see why I needed my iPad to be able to read about these amazing people. So with that said, now that we know who we have up here, let's start with asking some questions. The first oh, of course you know if you haven't already got this thing, you can ask your own questions here. Slido.com hashtag delve. So what does it mean to you to stay human in the age of AI? And whoever wants to go first, make.

[Ben Gutman]

A jump for answering well, so, you know, we're in the brand and advertising track. And when I think about what effective advertising is, it's something that connects to the, our innate human desires or innate human needs and wants. And I think that when we look at how AI plays into that, we're, we're often thinking about how we use AI as a tool to just get the job done. We got to get a whole content calendar out there. We have to get the blog post up, we have to get the SEO done, we have to get this stuff. But I worry that when we do too much of that, we're sacrificing our ability to do what is actually the real goal of things. It's not just the thing that's right in front of us, but the thing that's actually what we're trying to achieve, which to make that connection between us, between our audience, to fulfill their wants and needs. That's part of the way I look at that little, little bit of optimism in terms of how it helps us do our job, a little bit of hesitation in terms of it replacing our job.

[Joanna Lepore]

I'll also add one thing. I mean, in relation to marketing and branding, I think AI has been misunderstood. People think of it as doing automation. I think people ignore how simple and stupid the algorithms are. And they're just amplification algorithms. They're just technically, we call them stochastic parrots. They will parrot back what they have seen before. They don't even understand the words they're saying. They don't even say words. They say tokens. An English word is about two or three tokens. So what is not being covered by all the big companies in AI is how much human is in the loop informing these models and how much guidance that AI needs to get from humans all the time. OpenAI avoided talking about this even though they spent billions of dollars on it, both in curation and in humanly challenging the models to make them sound smart enough. They started talking about it big time. When Deep Seq came out, they started crying foul. You cheated. You used all that human intervention we did for free because you used our models to distill. They used the OpenAI models as essentially a teacher in a distillation process. So human in the loop is a big deal in communicating. And I think we should all be thinking about AI as a tool, not as AI as the primary means for communicating or letting our automation take over our marketing, our branding and all of that. I don't think that's possible today actually, effectively at least.

[Oussama Fayed]

Joe, anything to add?

[Ben Gutman]

Maybe I'll add a lens that's slightly less positive, maybe building on some of what you were touching on there. So when you speak to marketers and branding people about AI, they tend to get generally quite excited for very good reasons, because we're always as marketers looking for new ways to build those connections with our consumers, to be creative, to be differentiated, to stand out from the competition, competition to stay relevant. And that's all valid, but the risk that we run is that we're at the moment using a lot of AI in branding and marketing to outsource ideas, outsource gathering the intel, outsource creating campaigns, outsource doing the thinking. And some of the recent studies that we've seen are quite concerning in this space where AI is decreasing our ability to think critically, critically, it's decreasing our cognitive strengths and even our memory. So unfortunately, when it comes to anything in life, it's use it or lose it. And that's exactly what's happening. When we're continually getting kind of in love with this idea that AI can do so much for us, how do we make sure that we're not minimizing our own human strengths and capabilities in that process? And then I guess, you know, it's worthwhile pointing out that at the moment in branding and marketing, what we're predominantly using AI for is chat bots or the customer service that we're outsourcing to, you know, the AI. And what we've been seeing from research is that consumers are actually really disappointed with the outcome. So this is the most invested in area from a branding and marketing perspective right now, for good reason. I mean, I think we've all engaged with chat bots online, but the customer is telling us this is not a very good brand experience. And actually I'm going to go to the brand that's not using AI to converse with me. And again, if we think about the point that Ben made, if we're really trying to build relationships with our customers, in the end we kind of have to weigh up what kind of an experience we want them to have and also what kinds of skills are we building in ourselves that's not being outsourced to the AI? So maybe a slightly more pessimistic down out response to get us started.

I think That's a really good point. I mean, so quick question. How many people here have used AI to produce something that goes out into the universe, like a blog post, social media content? Okay. Most people. Looks like how many of you want to read something that was created by AI on a blog post or on social media? Okay. Like two people. So there's a little bit of a disconnect there. A good friend of mine, he was posting about all his work accomplishments and he said, oh, for this year, I want to do a lot more social media content. And so this is how I'm going to do it. I'm going to go to ChatGPT and I'm going to say, here's all my previous writing. Here's the stuff I want to talk about. And it makes me a content calendar. And now every time I see something from him on LinkedIn, I scroll right past it because I know it is exactly that kind of slop that's been put out there. And. And so I think what you just said, and I think what we just saw here is exactly that disconnect, which is all of us want to use it because it's a really sexy, you know, easy cheat code to get stuff done. But when we're the receiver, none of us want that. And I think that's the problem that we're going to see become more and more in the forefront in terms of marketing.

[Oussama Fayed]

Absolutely. And knowing that there are humans on the other side, I think being that point there, Right, Ben. That we're. We're needing to think about the person that is going to be reading it and why they would care. When you ask the right type of question, when you're thoughtful to why you're wanting to put something and it's a first draft that you can then edit is very different than, you know, oh, good news. I wrote that for me and it's ready to go. Because that is not usually the case. To add one additional element on here about what it means in staying human with AI, Really, I like to look at that internal communication and change element. I wrote a book on this about, you know, knowing what your employees need and they can't tell you and being able to understand that people don't just adopt change as easily as we think that they should, even if it is something that would be beneficial for them. And so understanding the way to communicate the value to those internal employees is a really big way to also look at this other side of the humanness. So we've got the customers who we want to find the content appealing and valuable and interact with the chat and think it's a good experience, experience in an up level for us as well as the humans that we're needing to be doing this, this work internally, even if we will be able to streamline processes a bit as we go through. So those are a couple things that we're really thinking about here. And so, you know, next big question here is, you know, do we think that AI is making us better or worse at communicating? I think we touched on this a little bit in that last round of questions, but anything to add for this point?

[Ben Gutman]

You know, Joe mentioned these, this kind of set of studies coming out about how much reducing our cognition sometimes to be relying on ChatGPT or equivalence. It reminds me of a group chat that I'm in with two friends that is basically just them arguing back and forth via chatgpt things that they've pasted back and forth. I think that it can go both ways and it's a really like kind of weasel the answer, but it can go both ways. Right. It can help you if you use it as a tool to improve the work that your, that is your idea. That is, you know, helping you get to the beginning of format forming something, helping you get from your first draft to a better written thing. But I think that when we use it as a replacement, I think that's when things fall apart.

I mean, I guess one thing to note is like we say AI, but generative AI is not all of AI.

[Joanna Lepore]

Right.

[Ben Gutman]

Osama speaks to this quite well and AI is not new. But that being said, you know, the way that we're seeing all the advanced models start to come out now, I mean, it's worth saying that this is improving and it will improve. So it's easy to be kind of critical of it right now and maybe point out its flaws and what's not working well. But like with any innovation that's seeing huge amount of investment and I mean R and D and AI, we're seeing, depending on the estimates that you look at $100 billion to $500 billion, which is huge. We're going to start to see more and more improvement in this space, which means that we will see it continue to evolve. And so we do need to be at the forefront of playing with it, experimenting with it, just, you know, with that caveat that it might not be perfect right now, but it will get better.

[Joanna Lepore]

We don't have very hard data on this, although Joe this morning sent us a link to a group at Northeastern University. Check it out on Northeastern Global News where they actually published a study saying that people are losing confidence in brands because of the AI communication, but taking it to human to human communication. My personal this is not backed up by data. My personal guess it's making us much worse, not better. We've become more effective at gathering information, quickly pretending we know something about some topic because you did a quick lookup. You know, this is kind of a next level of Google search, I guess, but it actually detracts. It doesn't help because you're talking to a human. It's a slower bandwidth, likely to be kind of less accurate and less complete. However, that's what we seek when we talk to people. We don't want to hear from an AI. We want to hear what that person thinks. And I think this is beginning to cloud it and make it less effective.

[Ben Gutman]

One of the use cases that you see in every big AI company's commercial is, oh, I have to write this email for my boss. You know, here's the, you know, sloppy version of it. Can you make it look nice? But then that email goes to the boss and then what are they going to do? They're going to use AI to summarize it and make it easier for them to read. And I think that you're right. We're just putting this kind of unneeded and unnatural kind of filter in between. Because it also reminds me of an early, an early instance where AI kind of broke through when Facebook, I believe they had, Facebook had some bots negotiating ad rates and they ended up developing their own language because the language that we would use to do it, English or whatever, wasn't good enough. It wasn't efficient enough for it. And so they had their own kind of language that they were using between each other that was indecipherable for us for that. And so it's a version of that, just extrapolated. And you know, I think that it really in many ways probably puts up more barriers than it breaks down.

[Melina Palmer]

All right, on that note.

[Oussama Fayed]

No. Okay, so as we continue here, really good points, of course, about things to be keeping in mind as we are looking to be, of course, more human in the communication. It's not all negative, but there are important things to be keeping in mind. And I do want to transition a little bit then as we think about for, you know, these brands and using AI and we're seeing some, some negatives, some positives here. Many brands are leaning into AI, some more publicly than others. There were ads from Coca Cola, you know, Saying this is produced by AI. Toys R Us did one as well. And often, you know, we have customers and consumers saying they are. Creepy is a word that has been used quite a bit with some of these. Apple has had some negative backlash on ad attempts as well. Skechers like multiple things that haven't quite landed there. And so, you know, there's a. They say there's no such thing as bad publicity, but I'd say there's a fine line right as we look at this. So question for those who are here representing brands. Is it better to be an early adopter to get out ahead of it? Because we're talking about those brands now and showing that they tried something, or would it be better for brands to wait until things get to that point you were talking about, Joe, that as it's getting better to be a little bit safer, which way do we think they should go?

[Ben Gutman]

Well, I mean, I think it's pretty clear that there are pretty significant brand risks when it comes to using and leveraging AI. Some of the more interesting studies are telling us that companies who are playing with AI in branding and communications, and at the moment, McKinsey estimates this is about 1.3 to 3.1 billion, which is still pretty small and growing. But those companies that are investing in it see brand risk as in the potential that it will impact your reputation negatively. That, you know, the. That it will say something that is not true, that puts your brand at risk or even from a brand image perspective is one of the biggest, if not the biggest, risk for the company right now. And yet the same CEOs are saying that they don't have any measures to account for that. They don't have any kind of resourcing or any kind of procedure put against that to be able to protect the company and the brand. So I think it just depends. I think it's eyes wide open, going in and really recognizing that this is a real thing that not only might happen, but will happen. And I mean, is there anything more important for a marketer or a brand leader than protecting the reputation of your brand? It will absolutely make or break your business, not just your brand, but your entire business. So measures have to be put in place.

Absolutely. And so when I look at it, some of those examples you said the Skechers one was something making the rounds. There's at my subway stop in Queens, there was a billboard for sketchers that was very obviously just somebody typed in make a shoe advertisement in AI. And it's the kind of thing where they got the job done, they saved however many dollars they're going to spend on a designer. But to your point about the brand risk is everybody looks at that and goes, what am I looking at here? And it doesn't even have shoes that Skechers would sell as part of it. The way most people may have seen that ad is because it's gone semi viral in terms of people making fun of it. But I think also when I get into the. The use of it. Does anybody know Pentagram, the design agency? Pentagram is maybe the most prestigious design agency in the world. And they recently did a project for the outgoing Biden administration that they did several hundred different illustrations, all generated by AI. And it was something that. It was one of the first times that you saw a company of that type of prestige, of that type of reputation use it. And they were very proud about it. They went out and they bragged about it on Instagram, everywhere else. And then the entire design community kind of went, well, what the hell? This is something that is going to jeopardize your jobs or something that is copyright infringement, whatever it is. And so it's interesting to see that we notice these ones because people are. Either they're visibly AI and we're like, that looks a little cringey. Or they're bragging it from the roof, shouting it from the rooftop, saying, we used AI for this. But it makes you also think a little bit, how many times have these different tools been used and we haven't noticed them and they've kind of just slipped by as something that plausibly was by a human designer or whatnot? The brand risk is there, but it's also one of those things that's only there if it gets caught in a way.

[Oussama Fayed]

A question that did come in through our slido hashtag delve, you know, on here, saying AI is very biased. Do you think brands have an awareness of this or do marketing professionals go through training on this? It kind of ties into their saying responsible and ethical AI use, which I know we're talking about later. I would start with you, Jo, if that works.

[Ben Gutman]

Yeah, definitely. So just a recap that I've recently come from working inside McDonald's and before that at Mars Wrigley. So, yes, absolutely. Brand leaders are aware, creative agencies, strategy agencies that are advising marketers are absolutely aware that there is a bias there. And there is sometimes an overcorrection as well on the bias, which we've seen also doesn't garner very positive results. So how do you mitigate that? How do you account for that. We did a piece of work at McDonald's from a foresight perspective. So foresight being anticipating what's coming in the future and kind of facing into the areas where we have less certainty, less data and knowledge. One of the biggest things that we flagged was this risk around ethics or unethical use of AI or a perpetuation of bias. And I guess, you know, it's been said before, but it's worth saying again that the bias exists because bias exists inside of human beings. The systems are built on data that we have created. It's picking the data that we choose. And the data is imperfect because we are imperfect as human beings.

[Joanna Lepore]

My comment here would be I agree that some brand managers are sophisticated enough and aware from what we see, a lot of them are not. They don't understand how deep these biases can run and they don't understand kind of the unintended consequences that come out when you design an algorithm, you know, designed to do some good, designed to have a one effect, but it has another because the data you fed it is not representative of your whole customer population or everything you care to kind of represent. We have at the Institute for Experiential AI, we have an AI responsible AI practice it's job. We created that practice because we needed a mechanism to engage with companies to understand. You know, a lot of people talk about responsible AI. Nobody. You know, you look for definitions, you get different definitions and all of that. So to understand it, we figured the best way to do it is to actually engage in projects and figure out like a, what should we be teaching the future. Engineers, computer scientists, social workers, what have you about this? And then figure out what are systematic approaches to detecting and mitigating that bias. And the bias comes by definition. Any data set is a snapshot of the world which will have biases in it that whoever collected the data didn't quite understand. And there's so many examples of that, including facial recognition, not working as well for darker skinned human beings and so forth. Not intended, but just never thought about it. They developed it on huge numbers of images of a lot of white people. So these are things that we need to be aware of. People care about doing responsible AI. And there are mechanisms and algorithms where you can audit systematically and find out where the biases may be and what the risks are. So risk assessment is a big deal for us. And I think the good news is it's doable, but you just have to bother with it and make sure that you do it as part of your practice.

[Oussama Fayed]

Absolutely. Thank you both. Thanks everyone for answering those questions. So we have another kind of fun, I guess, note in here that leads into the next kind of slido poll that we already have live and open, which says that every description at south by Southwest contains the word delve. Why don't brands care and filter these words out or learn to produce better writing? So the poll that we have in there now on the slido is, what are other telltale signs that you look at something, you go, that's AI when you're reading it. When you look at it, there's this interesting piece just from the behavioral science side of things. You know, so much of what we do every day is processed on what I just call subconscious kind of level, right? That we are just making very, very quick decisions and being able to. That's where when we see someone else put something out there and you can feel something is off, but when it's your own writing, you're using this more conscious brain and you're thinking about it in a different way, and you just don't read it the way that you would when you are the consumer. And so you don't get out of your own way enough to be able to do some of these things. And so being more thoughtful to that. There was another question here that was, you know, kind of asking about, you know, just why, why don't we. We do these things more? Why do we just let it go? And it has to do with, I mean, we got a lot going on, right? We're busy and trying to do all these things. And so you think it's gonna be good, it's as good as it could be or better than the nothing that you would have put out. And so you just kind of let those things slide. And so sometimes doing less, even when you are using the AI is valuable there panel, do you have some other quick things that are very obvious, AI, like, what would you add?

[Ben Gutman]

I mean, I think Dell's a great one, obviously, you know, innovative, emerging. There's all these different kind of fluffy adjectives that go out there. And to your point, these are the reason we keep these in is because they're easy, right? It's easy for us to just take the output and kind of copy and paste it into something. And it reminds me of a quote that's misattributed to Mark Twain always, but I wrote you a long letter because I didn't have time to write you a short one. It's easier for us to put all the words up and to spit everything out than it is for us to do the editing, to do the curation, to make something easy, make something fluent, to make something simple. And I think we have a misplaced confidence in the idea that big words make us sound smart. But in study after study, when all sorts of different type of researchers look at this, big words do quite the opposite. They make us sound that we're putting up facades that we don't know we're talking about. And I think AI is only kind of accelerating that.

[Joanna Lepore]

There's a paper published maybe four or five months ago by Professor Byron Wallace, who's part of our Institute for Experiential AI at Northeastern. And this should scare you. If you use AI, the ability to detect, just by doing very simple statistical analysis of the text, the ability to detect that it came from an AI is amazingly accurate. Right. And surprisingly simple. So keep in, you know, there's no free lunch. Basically, if you use a machine, a robot to help you produce something, it will sound robotic. It will sound, even if to a human it doesn't sound that way, the signature is easily detectable. So something to keep in mind. And it's a good paper to read. It's a very quick and simple read. It's a technical paper, but it's a good one. Wallace, Byron Wallace.

[Oussama Fayed]

All right, well, we have some others. As someone who is a big fan of the EM dash, right? Come on, man. People use EM dashes. That's that long dash for those that aren't writers. So apparently that's out now. Missing fingers, bad hands, overly flowery language, lots of words, but not a lot of ideas. Bullet points with the conclusion of a very short paragraph. Too perfect. Which gets back to one of the points we were making before. Right. We want human interactions with human people and those are not always perfect as we are going to be having these sorts of conversations.

[Ben Gutman]

So boring as well. Can I add that?

[Oussama Fayed]

Yeah.

[Ben Gutman]

AI content tends to be very boring.

It's sixth graders trying to fill out the word count on an essay a lot of times I saw somebody once use the tool when they make a prompt to just say at the end, no yapping. And apparently it quite improves it.

[Oussama Fayed]

Right. I've had to do for mine where when I'm trying to do a post or something to like, I'm not a 16 year old influencer. Can you please write this in a different way? That would work for me. So we talked about some pitfalls here. Anybody really getting it right that we should be looking to any, any wins out there?

[Joanna Lepore]

Yet, yeah, I mean I. There's several great examples. My favorite one these days, probably very relevant to this audience, is what Adobe did with Firefly, which is now part of Photoshop. Here's what they did, right? They used their own data, 40 years worth of image libraries that they had nothing from the outside, which enabled them to completely indemnify the users against any IP rights, any transgressions. And then they basically chose a very natural way as an interface to embed it. Right? You could be in Photoshop, you could pick a background and say, fill with a background that looks like this. Or you could click somewhere and say, hey, add a classic car in the middle of this desert scene and add a reflecting pool in front of it. And hey, add a cloud that looks like the classic car. That's a real example, takes about 30 seconds and it looks really good. So to me, just thinking about those tenets, which is where does the data come from? Do you have full rights to it? Can you indemnify the users and can you embed it in an intuitive, easy to use interface is absolutely kind of a good example. And I don't work for Adobe, although I like their CEO.

[Ben Gutman]

Yeah, I think that's a really great example. And often when we talk about examples in marketing, it's the creative execution or the communications and kind of what's visible to the public. But a huge part of marketing is designing great products. And I think in product development we see some tremendous advancements being used through AI because obviously it can sort through so many different, different combinations in the formula much, much more quickly and efficiently than what we can and in essence more creatively. So I think a great company doing that well is NotCo and they have their Giuseppe AI platform and they are creating lots of plant based formulations that allow us to, yeah, look at, look at existing problems in new ways. And that's what I get really excited about is, you know, we're thinking about sustainability. How do we replicate the experiences that we have now in a more responsible, ethical way. I love that application of AI And I also do not work for NOT co, but I really like their CEO.

[Oussama Fayed]

I would add as I guess on the design side too, as an individual creator owning my company, we use Canva a lot and there are some good options in there that are filling in better. When background remover in Canva came out, it changed my whole life and now you have the grab text and different things that you can do that does pretty good job with creating new images and things as well. There so, Osama, a lot of what you were talking about was a really good transition into, you know, talking about how we start using these tools. Right. So how should organizations be starting their AI journey? You know, what do they need to know? I know. So you were saying some things. Joe, do you want to kind of kick this one off?

[Ben Gutman]

Yeah, I think there's a really interesting conversation taking place right now about whether or not you need to do case studies in your business. Because the positive thing is that if you create a little bit of an innovation lab where you can really carefully and methodically test out a few different applications of AI in a really measured way. So obviously, really thinking about where you apply it and where you don't apply it, that can be really helpful. But the, I guess the counter argument that's taking place in that is that it kind of limits your. Your ability to really see the potential of the AI. So if you get people from across your business who put their hand up and say, I've got a great idea of how we can use this, or I just want to give this a try, saying no to that because you have the stringent set of case studies does limit the potential of your application of AI. So I think it needs to be a combination of both. You need to go in with a really measured plan of how you're going to apply it in the biggest bang for your buck ways. But then you also need to leave room for the employees in the business to really leverage it. And part of the reason for that is that we see from all of the studies that companies and CEOs are really excited about AI and they're expecting those returns in the next years, few, few years. But the employees feel really left out of that loop. They feel like they're not getting the training, the. The support, the capability, the access to the AI, generally speaking. And so I think it needs to be working hand in hand both the company objectives and upskilling employees.

[Joanna Lepore]

On starting the AI journey, I have maybe three things to say. You absolutely agree with Joe. You have to start it. It's not an option. I get asked the question all the time, will AI replace my job? And my answer is always, no. AI will not replace your job. But a human using AI will replace your job. The same holds for businesses. This gives you such an advantage in eliminating a lot of the expensive, robotic, repetitive stuff, which we have done an amazing job embedding in our knowledge economy. This helps counter it and helps you function faster with it. Now, the two things that I would say as advice on starting your journey, other than you absolutely must start it, because you got to practice the muscle and you got to figure out where it fits. Two things that most companies don't think about. Number one is data. By the way, all of this AI, all machine learning, which is generative AI, is a subset of machine learning is dependent on data. And most companies don't collect the data they need. And those who collect the data are collecting data for human consumption. So they're trying to aggregate it, summarize it, look for insights. And that's exactly the opposite of the type of data that's needed. Algorithms need completely different level data where you're basically capturing every little detail about everything. So capturing that data, including all the human interventions and corrections, is absolutely a must. And that's. You're not going to get that with ChatGPT or any of the public models. You will get that with small models that you can build. They're very narrow, they're specialized on the task, but then you can retrain them very quickly. ChatGPT, you know, ChatGPT 4, the last time was trained it was in 2023. Now there's a new version, 4.5. But to give you an example, the reason it hasn't been updated is because it's so darn expensive to update it that they can't afford it, right? So they have to work with snapshots. That is not the case. So collecting data is an absolute plus. And the second one is human intervention. If you use any of these models, giving it corrections, saying this is not right, do it this way, do it that way, that is absolutely very valuable ip, it captures the way you do stuff. It captures your processes, it captures your targets. You cannot do that with the public, very large models. And that is the other thing you need to think about. How do you create a solution that captures a lot of that? Correction, course correction, intervention, all of that that you do when you're dealing with AI models. And the last thing I'll say is no AI should be talking directly to a human or making decisions of import, because we don't know much about how the large language models work. I spend a lot of my time on that. But we know two things for sure. If you ask a large language model about something it hasn't seen before, number one, we know this, it does not know that it hasn't seen it before. Number two, it will actually produce an answer and it will sound very confident in the answer. And number three, with very high probability, that answer is total junk.

[Oussama Fayed]

And that's how it stays human.

[Ben Gutman]

And the only thing I would add is this is great, is don't fire the humans.

[Joanna Lepore]

Right.

[Ben Gutman]

Because, you know, to all those points, this is only something that is an accelerant on your own human capital. It's something, you know, the same way, you know you're not going to fire all your humans because you got to use a word processor, you got to use a spreadsheet, or you got to use Photoshop. Those will take the people that are this talented and make them, you know, make them this good to do this over here, raise the floor, raise the ceiling. At the same time, all of those things are important. And especially when you look at how kind of like where the creative pendulum tends to swing in marketing, advertising and culture is that the things that are difficult and expensive tend to be the things that end up being valued a little bit. So like when photography was very expensive and rare, we had a lot of illustrations. When photography got cheaper and more accessible, we moved back to, I'm sorry, other way around. When, when, when, when the corporate Memphis style. Have you seen this ever? It's kind of like the cartoon characters that we all had in the 2010s on our website. All these big companies went out and did that because it was very easy to do photography. And then we swing back this way and say, the stuff that a human creative can do is something that is going to be more valued and different. And I don't know if that. What that becomes is that something where, you know, it's an analog thing. I mean, when you look at how high school teachers are teaching now, they're saying, well, we have to do a handwritten essay for something or presentation in class. So the things that can't be automated are the things that end up being valuable. And so don't replace the humans just because you can automate some parts of their job.

Yeah. And I guess just to recognize that we are seeing that, though, we are seeing people being taken out of their jobs and replaced with AI not just in the news, but I'm sure in conversations that you have had. And what we know is that training somebody to do a new job takes 12 months on average. So that's a lot of lost knowledge and talent. So I guess, you know, a reminder that it's. It's not just a risk that I won't take your job, but someone will may fire somebody and put AI in that place. And that's a huge loss of the. The human intel and the talent pool. That is a very real risk.

[Oussama Fayed]

Well, this comes to another one of the questions coming in on the slido here, which is saying, you know, how do we inform or convince executive management at our companies that AI is not necessarily the best way to go to make things easier? And there were some other questions having to do with that. I think that Sama touched on this a little bit as well. And for me, from the behavioral communication side, one tip that I would give to start off here is to know that in the way that where AI might not replace your job, someone who's using it very well could. And so if you can get ahead of it and understand how you can find the value proposition of what it is that the AI can take away from the part of your job that you hate doing anyway, like, what's a thing that is taking up so much of your time in this? Maybe it's, you know, before you can even get to the point of being able to do the stuff you really like, or you never even have time for it because you're having to process this particular report, or you're having to sift through some other information, or there's something that you're doing that takes you a very long time but could be done by the AI very quickly and then opens up an opportunity for what you as a human can do in a better way, that the entire department can be much more efficient, effective if you can get ahead of it and present this to the leaders around you and say, this is how we can be this much more efficient, effective, and when someone leaves, we won't have to replace them. Or look at, we can do double the work that we're doing now if we're able to have the AI work on this thing so that the humans can do this thing. And we're uniquely able to do this because we can ask these questions and we have this background and this is how it will save the company money, or this is how it's going to help us to be more effective, or this is how we can track sales better than what we've been able to do before. We can run more ads on social media and see what's working and replace them more consistently and see how effective we are. And we can transition faster if you can speak their language of something they care about in a way that you can use the AI to help support the work you don't want to do anyway to make it so you like your job more and have better balance. There is a way for it to be more of a win, win, win. If you're thoughtful to that. And if you don't get ahead of it, they might pick something and implement it in a way that kind of sucks. So being thoughtful to that, that would be where my advice is. Anybody?

[Ben Gutman]

I'll just, I'll just say, to quote Bruce Springsteen, nobody wins unless everybody wins. And that's exactly how you got to look at it in corporate changes, that everybody's in it for their own thing. Everybody has their own initiatives. The executives, they want to look cutting edge. They want to keep up with their peers. They want to say, you know, in the news and the press release them using AI for this and that. And so if you're trying to, to get some sort of change through that institution or through their head individually, you have to exactly as you said, speak their language. What is their own motivation? And this is great kind of client services recommendations too, is like, everybody has their own motivations. You got to figure out what they are and how you can help them achieve their goals as part of what you want to do too.

[Joanna Lepore]

One of the quotes I really like says AI is not about replacing the human with a robot. AI is about taking the robot out of the human. And I think that's how we should be thinking about it.

[Oussama Fayed]

Perfect. And thank you. We still have, we have 19 questions that are in here, and I've been going through them as we've been going along the way. So thank you for everyone for being so engaged. We, of course, don't have time to get to everything that's in here, but we do have one with five up votes, which we'll do here, which says, what can we learn from the pushback and then adoption of other transformation tech like TV and Internet, to not fight AI adoption, but to help shape it for good.

[Ben Gutman]

Does anyone interest question?

[Oussama Fayed]

I know, and that's into where's all this going?

[Ben Gutman]

Congrats on a great question. Yeah, I mean, one thing that I'll say, and I'm sure the panel can build, is that we, we are obsessed with innovation as a species. And we think that innovation requires technology because technologies and science and engineering has fueled so much of our progress, particularly in the last 150 years, we've seen a huge exponential increase. And by the way, all of that has come with economic prosperity. And despite the fact that we are seeing a lot of, you know, disparity and division around the world, we are getting better, we're becoming healthier, wealthier, we're living longer. All everything's pointing in the right direction. And a lot of that has been fueled by all of these technological advances that allow us to live the very, very comfortable lives that at least we in this room live. Although it's worth noting, 40% of the global population don't have access to the Internet. So we are in a very privileged room right here. So one thing to go back is like, yes, technology will continue to be a huge part of our lives. Yes, AI is going to continue to grow and transform our lives, in part because we see transformative innovation of the past has really filtered through to multiple industries and sectors. That's how you get really, truly transformational technology. And that's exactly what's happening with AI. It's not just doing advancements in medicine or influencing how we buy media, or influencing, you know, the shitload of crappy Amazon books that are written by AI. It's. It's really filtering through to everything. And that's what it will continue to do. But just for a second, I want you to imagine a world that does not have technology, that does not have AI, where in the blink of an eye, all of that is switched off. What would your employer want you to do? What is that human element of critical thinking, of creative thinking, of imagination that you can bring, that doesn't rely on technology, that doesn't lean on outsourcing something to a task? We're obsessed with innovation, we're obsessed with technology, and we're obsessed with efficiency. And that is not a natural human state for us to be in. So I guess a provocation around the future. If yes, the future includes technology and AI, what is the human future that we want? Because AI and how we use it is a choice.

[Joanna Lepore]

Yeah, I would, I would say, look, AI is here to stay. For the simple reason which I mentioned before, we have created a knowledge economy that is just rife with robotic, repetitive droning tasks. Right? For good or for bad, that's what we've done. There's many estimates of how much would it be worth. But I'll just tell you, when I say knowledge economy in a country like the US, 51% or so of the US economy is a knowledge economy. My personal guess is that these AI technologies will result in anywhere between 50, 15 to 80% acceleration, not total automation, but acceleration of a lot of these tasks in the knowledge economy. So even if I'm wrong, I mean, 15 to 80% is huge. Even if I'm wrong, even if it's just 2%, 2% of that translates to 1% of the US economy. That's a number we cannot ignore. So it is here to stay. We should embrace it. The biggest discussions I've been having with the president of our university, Joseph Aun, who wrote a book, by the way, called Robot Proof Higher Education in the Age of AI, and he wrote that book seven years ago. Now, robot proof, I think, is the trick. His typical call to me is, Osama, what should we be teaching our students? Like, what's the value add? If a ChatGPT can pass a bachelor's degree and soon a master's degree exam and so forth, what are we teaching our people? And it's really about teaching people how to think in this new age and how to be robot proof. So that core idea is something to think about for your job, for your future growth and all of that. And there's plenty of room to be robot proof. You just have to understand that these tools are there to be used and they're just tools, just like all the other technological advances. Nothing scary about them.

[Ben Gutman]

And I'll just make one other point based off of both what you said is that even if OpenAI goes out of business tomorrow, which maybe even if the investment in these data centers from Microsoft pulls it down or Google spends too much money, even if all these companies go away, AI in terms of these different models, something like Deep Seq or these other models that can be run locally in your own computer, it exists. It exists in the same way that Super Mario Brothers exists. And you can play it no matter. You don't have to go out and rely on Nintendo existing anymore for you to have that software. Regardless of how we kind of settle in. And I always like to mind the hype cycle where things get overhyped and underhyped and back up and down. Even if we're kind of getting whiplash from this at some point, this software will always be there and you will find a use for it that hopefully won't be carrying the brand risk and the detrimental stuff that we're talking about. Hopefully we find a place of equilibrium that ends up being an accelerant and an enabler for us instead of a replacement.

[Melina Palmer]

Wonderful.

[Oussama Fayed]

Thank you. There's another question on here saying, asking if the Panel thinks that ChatGPT Gen AI can really help non English speakers get a level playing field. And that's some of what they're seeing with some different types of designs and designers. And I think that there's some interesting things, even with something like, you know, Duolingo is using a lot with AI that you can have conversations now like you're on the phone for simple type of language that comes up there. And then we also have different translation that can be done for podcasts and things that I think is really interesting. So any other thoughts on that?

[Ben Gutman]

The translation from these platforms is one of the use cases today that it's the best at.

[Oussama Fayed]

That is, though, a good time as we go ahead and start to close and to say, what is it that as you're looking forward, what are you most excited about? Or what would be a last thought that each of the panelists want to add. For those in the room, I'll start.

[Joanna Lepore]

With the super negative here. So I'm not, you know, there's a lot of people, very smart people, very respectable people who worry about the super intelligence and the aid, you know, AGI. And I don't. I think the current technology does not. Does not get us there. Right. And that's not a real threat to humanity. I think there's a very, very real threat to humanity that's hitting us right now. The ability of this AI to basically create a world in the digital universe we exist in, on your phones and online and so forth, where you cannot believe anything you read, anything you hear, anything you see. Because the ability to leverage the technology to create. If I wanted to create a bad story about Joe with a press of a few buttons, I can also not not only generate a thousand variations on that story, I can have a thousand websites with thousands of robotic followers to each website attesting to the fact that, oh, this is so real, etc. And I can create deep fakes, voice renderings, all that. And you're seeing that today. You're seeing it in very bad advertising, but you will be seeing the danger is the bad actors are much earlier adopters than kind of the good actors, because the good actors have to worry about things like the law and compliance and other things. And that's the world I worry about. If that turns our digital world upside down and that's complete disorientation for our existence out.

[Ben Gutman]

Leave with a thought about kind of the creative, you know, creative direction element here. So we, when we look at how AI works, as I think a bunch of us have said, it's an average machine. It gives you middle of the road results, gives you middle of the road thinking. It takes all the information out in the universe. All the images gives you the average of something. All the text gives the average of something. And I know that's a big simplification of it, but when you look at what we notice, what we pay attention to in advertising and other creative endeavors, we notice what's different? We notice what's on the edge case. It's the thing that stands out. It has contrast. It's something noticeable. And that's what is good marketing. That's what is good culture. And I think that that's the important thing to keep in mind here is that we can't rely on this to give us the weird stuff. We have to do the weird stuff ourself. And that's the thing that's going to move the needle.

Yeah, make more weird stuff. I like that. I guess one for me is that yes, we are a collective and a society. We're social beings. It's part of the reason why we humanize AI so much. But we are also individuals. So yes, get excited and scared about AI, but consider how you might use it. So just because the ability for you to use it in a certain way is there as a marketer or as an individual, you know, consider where you do want to apply it and where you don't want to apply it and make it, make it work for you and make it your own.

[Melina Palmer]

Wonderful.

[Oussama Fayed]

Thank you for those ending points, everyone and for the panel and all the questions. So many amazing questions. I'm so sorry we weren't able to get to them all, but thank you for engaging with us here again. So our contact information is up here. Usama, Joanna, Ben, and I'm Melina, as mentioned, Ben. And actually I both have book signings coming up here. His is at 4:30, mine at 4:45 at the bookstore at the convention center. Would love to be able to speak with anybody who wants to come by and just thank you again for all your questions and please do stay human out there, everyone. Thank you.

[Ben Gutman]

Thank you.

[Melina Palmer]

Thank you again to south by Southwest and the whole panel from our session. That's Ben, Joanna and Usama. And what about you? What got your brain buzzing in today's conversation? For me, I really enjoyed this panel. It was a fun challenge to moderate the discussion and manage the slido polls and questions rolling in from the audience. It is definitely different from just being a guest on a panel, that's for sure. It wasn't surprising to see that there was so much conversation about AI across South by Southwest. And then a few weeks later when I was honored to also be speaking at Social Media Marketing World, another fantastic event. I highly recommend AI is huge and it's here to stay. You need to keep up. And ironically, if you don't pay attention to that human element, people will likely be the downfall of your AI plan both on the customer side and the employee side. If you consider the humans, you're more likely to have success in your AI endeavors and behavioral economics can help you do just that. If you liked this session and would like to have me talk to your corporate team or conference about staying human in the age of AI, I do have a talk on this and I have done special versions of my talks on both customer and employee experience to focus on AI in the past for different sessions, events, things like that. So if it sounds like it could be a it, I would love to talk with you about it. Send an email to molina the brainy business.com to learn more or fill out the form on our website atthebrainybusiness.com and we would love to connect, learn more about what you got going on and see if we're a fit for each other. So what about you? How are you planning to use AI in your business? Come share it with me on social media. You'll find me as the Brainy Biz pretty much everywhere and as Melina Palmer on LinkedIn. There are links in the show notes to make it easy as well as links to my top related past episodes, books and more. It's all waiting for you in the app you're listening to and atthe rainy business.com 511and thank you again to Ben, Joanna, Usama and the south by Southwest team for making this episode possible today. It was a delight to chat with and learn from and with you. Join me Tuesday for another Brainy episode of the Brainy Business Podcast. It's going to be a lot of fun. You don't want to miss it. Until then, thanks again for listening and learning with me and remember to be thoughtful.

Thank you for listening to the Brainy Business Podcast. Molina offers virtual strategy sessions, workshops and other services to help businesses be more brain friendly. For more free resources, visit thebrainybusiness.com.