The AI Coach
AI meets human behavior in these fun and insightful conversations with Danielle Gopen, a top founder coach and advisor, and Paul Fung, an experienced AI founder. We blend AI breakthroughs with real-world insights on business strategy, industry disruptions, leadership, and psychology while cutting through the hype to ask the hard questions.
This is for decision-makers like founders and business leaders who want to harness AI's potential thoughtfully, exploring both the technological edge and human elements that drive long-term success. Join us to challenge the status quo, navigate the shifting landscape of an AI-powered world, and get new information worth sharing.
The AI Coach
Chicken or the Egg: Brand or Product?
In today's episode, we dive deep into the often overlooked yet crucial aspect of startup success: brand strategy. Many founders focus solely on product development and fundraising, neglecting their brand until it's too late. We'll explore why this approach can be detrimental and how intentional brand-building can set you apart in an increasingly competitive landscape. And of course, we'll explore how AI can help build a brand that resonates with your target audience.
We love feedback and questions! Please reach out:
LinkedIn
Episode Directory
Mentions of companies, products, or resources in this podcast or show notes are for informational purposes only and do not constitute endorsements; we assume no liability for any actions taken based on this content.
Hi Paul.
Speaker 1:Hello.
Speaker 2:How are you?
Speaker 1:I am good it's been a long week, but always looking forward to recording a new episode.
Speaker 2:Me too. Any updates that you want to share before we jump in.
Speaker 1:Updates In the AI world and I didn't think we'd plan on talking about this this week, but yeah, some new models dropped O1, I think, is interesting. Sam altman has said he thinks that they have reached agi level two. I guess there's like five levels of this right, so he thinks the reasoning is kind of at level two. So that's ai stuff this week. What are the updates? I think that you know that's probably the main one that happened this week. I'm sure there's a million other, as there are every week. There's always a million updates in AI, but that's the big one this week. What about you?
Speaker 2:What about me? Well, I will say about the Sam Altman interview, because I also saw that and it just made me think how did this one person become the face of AI? It's really crazy.
Speaker 1:Yeah, it is pretty crazy. I mean he has been well known for a while now through his work at Y Combinator, but I think you're right, like it is pretty crazy how these individuals can become the face of like a whole movement, a whole technology. I mean, I guess Elon is that in some ways with EVs and SpaceX and stuff, but yeah, sam really became a household name, which is wild.
Speaker 2:It's actually a good segue into what we were thinking about talking about today, which is the importance of brand, and so here you have these individuals who have such strong personal brands, and then how much that ties into the companies that they're involved with, and I feel like something that came out of the last two episodes that we did was this idea of okay, if AI is leveling the playing field in terms of development and engineering and AI is giving us so many more tools for sales and marketing and other things that are brand related, how do companies really stand out from their competition?
Speaker 2:So you mentioned before, like Twitter, for example, or now X, could have been replicated from a coding perspective within 24 hours if it were launched today Twitter X as we joke about brand you can't even agree on the name, but I think that also shows how strong the Twitter brand was, that even after it's been rebranded, people still call it that and I think the idea there is okay, so somebody else could come and replicate the interface, which is not that complicated. But the network effect, the brand let's call it the brand strength of the existing platform is what keeps everybody there, and you could say that's the case also for others like Instagram.
Speaker 1:But yeah, just thinking about brand, that's really been on my mind lately topic because, like you said, the bar for creating new software is so low, and so when the bar is very low, the ability for copycat apps to to emerge is is, you know, going to be faster than ever. The rate at which they're going to emerge is going to be faster than ever. And so, like in the past, you know there have been some startup. You know founders and kind of vcs who say you know there have been some startup. You know founders and kind of VCs who say you know the only moat is speed of execution, and typically what they mean by that is, you know, speed of shipping, right, like, how fast are you shipping new products and services?
Speaker 1:There's, obviously, execution involves a lot of things distribution, go to market, etc. But from a technical perspective, oftentimes I mean you outship your competitors. Can you ship the features that your customers need faster than they can? And so, when everyone's using now AI to increase their pace of shipping new products, you can't rely on just product alone or shipping speed alone, and so I think distribution and brand become even more important. I mean, you already are seeing this with the latest batch of startups, right?
Speaker 1:So, like in our cohort of companies that raised money last year for every kind of category or type of product. There's like 50 companies in the space. I mean that's a little bit of an exaggeration, but not that much of an exaggeration. And that's because it is just so much easier to build software now, especially on top of LLMs, that it's very easy to just like create new companies and copy the latest thing that you saw last week. That was really interesting. That's gaining traction and so in that world we have to find a way to differentiate ourselves, and oftentimes that way is positioning and branding.
Speaker 2:Yes, and I think I mean what you're saying is exactly the key point. I think a lot of tech founders don't think of brand. First right, they think of their product and then, arguably, b2c companies tend to be more brand oriented because their market of consumers care about brand and they need to establish a lot of trust with those consumers for them to be incentivized to buy. Actually, what we should do, maybe, is how do we define brand? I feel like it'd be interesting for us to even just say what does brand mean to us? Yes, there's a definition of brand and brand strategy and all that, but what do we really mean when we say brand?
Speaker 1:I feel like I'm going to come off as a terrible MBA student here. I mean, we both obviously met getting our MBAs and from that we should have a really solid I think, academic version of the definition of brand. But I think you mean, like what's the less academic version? Like what does it really mean at the end of the day?
Speaker 2:Yes.
Speaker 1:Would you like to go first? Because marketing was, for sure, my weakest suit.
Speaker 2:See classic tech founder.
Speaker 1:Yeah, exactly.
Speaker 2:So for me, when we talk about brand, what I really mean by that is a few things. One, it is the experience that your customer has at every touch point with what you're offering your brand. So anytime they see an email come through that has your logo, anytime they see a LinkedIn post with your logo, anytime they see the actual product that you're selling whether that be hardware or software or services or a tangible product they know what to expect from that experience and they identify characteristics to that experience. Obviously, I think brands in general care about being credible and high quality. But if, for some reason, your brand is focused on being low price, low quality, because the idea is that people don't care for that particular item and they're willing to replace it every two weeks because it serves a need during that time, even though it's low quality, then great, that's that brand.
Speaker 2:And so I feel like it's really about the experience a consumer has with your company name, with your product. And then, on top of that, what I think about from our worlds is what does brand really consist of? It consists of trust, so that if you're delivering something to that other party, they trust that it will be what they expect it to be. They trust that you are legitimate, your company is legitimate. And then the association of, again, quality and price and things like that I think that's.
Speaker 1:I think that is a great definition. The first thing that came to mind is is an outside of tech example of. I was like looking at some flights this week for a trip I need to take to the west coast and I was using kayak to search for flights and like spirit and frontier showed up and like their brand, I knew, I just knew you were gonna say spirit.
Speaker 1:Yeah, I don't know how, but I well, they're unabashedly low cost, right, and and that's their brand, right. They're not even trying to attempt quality. Yeah, exactly, and they're not. They're okay with that. Actually, even more so frontier's website actually said no, we don't provide wi-fi, and the reason we don't provide wi-fi is because it allows us to give you a lower price and so so they're not even trying to pretend they're a high quality brand that has any bells and whistles. They're trying to say we have no bells and whistles and basically, the thing we're allowed to offer you because we have no bells and whistles, is the low price.
Speaker 2:And, by the way, their flights are full. Yeah, exactly. So there are people who know that about their brand and they sign up. I was going to say they get on board, pun intended. Yeah, they like that, they want to be interacting with that brand and they appreciate the transparency. Right, like you're saying, they're not trying to be something they're not. They're very clear on what their brand is.
Speaker 1:Yeah, and I really like how you define it as kind of the expectation that your users or your customers have at every touchpoint, because I think that really is where the rubber meets the road for a brand right Like you could have you could put out a really shiny brand with like a really cool logo and like maybe some really cool ads, but if that totally falls flat when your users actually interact with your company or with your product, then your brand, your reputation, is not actually going to survive, right Like you'll actually get the brand or reputation of kind of being like a snake oil salesman, right, like, oh, they say one thing but do another thing and that's not what you want at all. Yeah, from a tech perspective I mean from a startup's perspective branding matters quite a bit in a number of ways. I mean the first and foremost one is obviously users and customers, and so I like what you said about their interactions. Oftentimes, on the go-to-market side or the sales side of tech companies and startups, we try to compete on being the most responsive salespeople, right? Or the most responsive customer success teams to make sure that when people interact with our company oftentimes especially the small of the startup we know that our users are taking a little bit of a risk or they're like an early adopter. When they're using our product, they know it's not fully featured and so the trade-off they get is really personal experience, right? So when we're working with our design partners, we might not have all the features they need today, but we're giving them, we're building, features aligned with some of their personal needs, because they're some of our earliest customers, and so in that way, you know our brand.
Speaker 1:Our reputation is that we're going to be really close partners with you, much closer than a larger tech company who won't be as responsive to your needs. And the other thing that I often say to some of the teams I've managed in the past is, especially in this world of AI, machine learning, something that we understand well, but oftentimes if your users are, let's say, business users, they don't understand it as much. We want to be seen as trusted advisors, right. We want to be the ones educating them on how to think about the problem correctly, and hopefully that aligns with our solution being the right solution for their problem. But even if we're not the right solution, for some reason we've left them with. They've learned something from interacting with us. They've learned how to think about the problem they're looking to solve. So a lot of times in tech for startups, you're trying to be a trusted advisor to your users, to educate them on the space that you're innovating in.
Speaker 2:I love that point because I think what you're also saying there which is something that has also been on my mind lately after doing our preparation for the LA Tech Week panel is helping companies really find their unique value proposition, and so it's like you're selling this, but what are you really selling? And I think what you're saying is, yes, you're selling the features that narrowed AI can offer, but what are you really selling? You're really selling where your trusted advisor, where are the people you come to, or the company that you come to when you're kind of a little bit stuck and aren't sure how to figure things out in this world.
Speaker 1:Yeah, especially in the early days, it's just like so important, right, because we don't have all the features that a bigger company is going to have, right, and so we are trying to be branded as we as long as as well as the rest of the other companies in our cohort, like I said, that have raised money in the last year we're all trying to guide users, customers, enterprises, small businesses through this new world of AI, generative AI and large language models, right, and it's actually quite a murky world for the outsiders, right. It's like this crazy new technology. People are afraid of that, people are excited about that, all of the above, right. And so, to the extent that we can help our users navigate that world whether they're business users, technical users, etc. I think that would be a branding success on our side if people see us as helping navigate them through that.
Speaker 2:Yeah, because I'm just thinking that even on the developer side, where they are tech savvy, they do have engineering backgrounds and obviously are very excited about AI there probably is also this maybe fear of getting it wrong and not using the right AI model for what they need not getting the best quality code when someone else out there is getting the best quality code and they don't even realize that there is that discrepancy because they didn't know exactly where to look to begin with. And so I feel like you help alleviate those problems and pain points by saying, hey, you can be reassured that when you're using our product, you are getting it right.
Speaker 1:Yeah, in some ways our products and we talked about it two episodes, what our product does. But just for anyone who didn't hear that, our product is helping optimize prompts for accuracy and cost and other reasons. But when you optimize them for accuracy and cost and other reasons, but when you optimize them for accuracy, in some ways we're trying to bring what we would call determinism to a non-deterministic world. Right and so like and kind of. I'll repeat the definitions of that. You know, gen AI, llms are non-deterministic. There's a little bit of randomness to the responses that come out of them. And what's funny is actually I was talking to somebody about this earlier this week A lot of software developers they hate prompt engineering specifically because it's not deterministic.
Speaker 1:Everything you do in software engineering is I write a function, I give it this input. Every single time it returns a similar output, right, and so this like very non-deterministic world of randomness, of like, well, sometimes I put in this input and I got this one output. Sometimes I put in this input, I get a different output. It's a very uncomfortable world for some people who like very deterministic software engineering and so in some ways we're helping bring more determinism to a non-deterministic world.
Speaker 2:I love that and I think, even as we're talking about this, this is a representation of the types of conversations that founding teams should be having as they think about brand. So, who are our customers? What are their pain points? What do they need? How do we solve that? Yes, everybody knows that.
Speaker 2:Those are the things that you should be thinking about, but then, on top of that, how do we become known as blank? And then that is what feeds into what your brand is, so the characteristics that you want to be known as out in the world. Yeah, I mean, this is just a perfect example of how these conversations can go. And then, on top of that brand is really a funny thing when you think about it. So the reason you see all these websites who have other companies' logos as social proof and credibility is because this is how much brand matters. So they're saying, hey, you should trust our brand and we associate with these other brands and that's how we help strengthen our brand back to us, but it's only because these other brands mean something. And so I'm just thinking, even from the fact that you're VC funded for your early users, that probably helps your brand, because early users are more willing to trust this product that's been delivered to them, that isn't fully fleshed out, because there are names out there who have deemed it credible. Does that make sense?
Speaker 1:Oh, yeah, totally. I mean, you'll see on a lot of startup websites and we don't have it on ours because we put up our site for our launch I don't want to say fairly quickly, we just haven't fully fleshed it out yet but yeah, a lot of startup websites. You will see their investors on the site because if they don't have a lot of customers yet, they can associate themselves with their investors, who are often the highest branded, highest quality or highest trusted kind of entity that they're associated with. And then I would say the most prominent example of that is actually any YC company, right. So YC actually has like a little badge that you can put on your website if you're a YC company, right. And so if anybody in the tech world goes to your website and they see it, if they see a YC company, they immediately think to themselves oh, these guys must actually be fairly serious or good at what they do if they got accepted into YC, and so I think that's.
Speaker 2:Which, by the way, speaks to the strength of the YC brand.
Speaker 1:Oh, totally, 100% right.
Speaker 2:And YC for anyone who is not familiar with that brand. Y Combinator.
Speaker 1:Yes, yes, yes. Why C for anyone who is not familiar with that brand? Why Combinator? Yes, yes, yes, and I think to your point earlier about this is something that founders should be thinking about earlier and earlier I do think it has been these days starting to get more drilled into our mind that there are some founders who will say the only thing that matters is distribution, or distribution is more important than product, even.
Speaker 1:Right, and there's like a little bit of I don't wanna say attention there, but like a debate that goes on in the tech community, and when I think about brand, I think about its impact on distribution, right, and so a lot of things these days, especially with social media, especially with influencers having more and more impact on people's buying decisions. You know it's funny years ago, influencers were thought to really only move the needle on consumer buying decisions, right? So like, oh, I see this influencer that has this purse or this water bottle or something like that and so, or these shoes, and so I'm gonna buy these shoes or whatever. But now more and more, influencers have actually quite an impact on B2B SaaS, like enterprise decision-making, right, you see, like-.
Speaker 1:Especially on X by the way, exactly On X and on LinkedIn, you see these influencers saying, hey, here's the products I use, here's the SaaS products that help me run my business. And then everyone else says, oh, I should be using that SaaS product. And so when I think about brand, obviously having a strong brand, being a trusted brand, leads to word of mouth, and that word of mouth is distribution. Right, that's how people find out about you, that's how people do discovery, that's how people enter your buying funnel. And so, for me, even if a founder was like, oh, why should we care about our brand? It's like well, do you care about your distribution? Because your brand is very tightly aligned to your distribution.
Speaker 2:I love that. It's a very important point and I think about that a lot too, and I think about we talked last time about the AI-enabled outbound marketing. So email and LinkedIn and I think if you're sending messages to all these people that are essentially cold messages yes, they can call them warm outbound because you have a mutual connection or something like that but they're basically cold messages. If you don't have a very strong brand, those messages mean nothing. They're essentially spam, unless that message is coming from a brand that I already recognize and or have some association with.
Speaker 2:I do not care at all about what you're selling, even if it actually is the best possible product for me. I don't have the time and energy to sift through. What is this company? Are they credible? Who's behind them? All of that stuff, and so it just becomes noise, and so I feel like being able to establish this brand, maybe even before distribution, which it might be a little bit of a chicken and egg situation, but if you have an early test group that you can use before distribution which it might be a little bit of a chicken and egg situation but if you have an early test group that you can use for distribution and they help make your brand really strong. They become essentially brand evangelists. That's one way to think about it, but yes, depending on the industry you're in and the product that you have try to do massive outreach and scale distribution before you have a strong brand is really worthless in my mind.
Speaker 1:Yeah, and I think I mean this is how you know, when we think about sales funnels for our startup, you know we often think of the marketing activities we're doing. It's like top of funnel, you know, middle of funnel, bottom of funnel type of activities, right. And so the top of funnel activities we're doing is just like branding stuff, right. So it's we're making sure to do our LinkedIn posts. We know that when we cold email someone, we know that they might not respond or even click the email until the third or fourth time. They see it right, because what we want them to do is we kind of want them to see us on LinkedIn and then also get an email from us, and then their friend maybe also saw us on LinkedIn and says, hey, have you heard about this? This is kind of interesting. And at that point is when they say, oh, maybe I've heard about this three times now, maybe this is worth checking out, right, and so I think you're right.
Speaker 1:In this world, that is, especially in the world of bringing it back to AI, the world of AI tools, which is so, so noisy right now, how do you stick out from? The rest is you need to have a strong brand that has people talking about you. You need to be putting out interesting content on LinkedIn. You can't just put out kind of thoughtless content. Linkedin's algorithm actually we've been looking at this because we want to do some more advertising on LinkedIn they prioritize more thoughtful content that people engage with, and so all of that relates to your brand is like what is the quality of the content you put out? What is the quality of the interactions with your customers, so that when somebody does interact with you, they go tell their friend hey, like I did talk to, you know these folks at this company and you know they're really smart. Maybe you should talk to them too. So it's just like brand is touched on in every single way. You know throughout of how you represent your company across the web and in your interactions.
Speaker 2:So my first question is what is LinkedIn deemed to be thoughtful content? I'm curious.
Speaker 1:I don't know exactly how their algorithm works, but from my research, I think both LinkedIn and other platforms are trying to, or starting to, prioritize what is the word for it, basically the time you spend engaging with a piece of content. There's a word for it that I'm forgetting right now, but basically, like if I think it is just like engagement.
Speaker 1:Yeah, there's another, there's another term for it. But basically, as you're scrolling through your linkedin, if you pause to read a post and actually like watch the video there, I think they prioritize that higher because they're like oh yes, I know, wait.
Speaker 2:Did you see the mr beast document that got?
Speaker 1:I haven't seen the full thing yet. I I saw some headlines around the summarization. Do you want me to read it?
Speaker 2:Okay, you should read it because it is fascinating. Speaking of brand, so Mr Beast is a brand, yeah, yeah.
Speaker 1:Incredible brand.
Speaker 2:Mr Beast is Jimmy, one person who has built this mega brand around him, and so this document is basically an onboarding document for new employees. And so he says I can't be everywhere at once anymore and the early team had direct access to me so that I could really be a part of every decision that was being made, but that's just not feasible anymore. So here is basically my stream of consciousness for how this brand stays strong and continues to be the best in the world, and this is what you, as the employee, need to think about and need to do. But in that document he mentions a few metrics. I mean, you choose specific metrics, but LinkedIn might use the same ones that are around what you're saying. This document is just fascinating. I really recommend reading it.
Speaker 2:And then, for companies who are out there hot tip for any founders who are out there who read this document and then think, okay, yes, this is applicable to Mr Beast. If I think about this for my own company, what would I write? Do it, go and write it. It doesn't have to be perfect. You'll read his. It's not perfect, it's not grammatically correct, it doesn't matter.
Speaker 2:It's also in his voice, which I think is actually the important part, right, because that's part of the brand. And so I think that goes back to even internal documents, our branding documents. All of that comes together to say who are we and how do we operate and how should we be seen internally and externally, and he makes a big point in that document. If you, as a new employee, don't agree with these things, just leave now because you won't find success here. And I think that a lot of people hear that and they think, oh wow, that's harsh or, you know, just kind of have a negative reaction to it. But I think, especially for a startup company, that that needs to be said sooner rather than later Because, again, if you think about your brand, you want everybody who's part of your team to be on board with this brand and with this mission and vision and be pushing forward together, collectively.
Speaker 1:Yeah, I couldn't agree more, and actually we've been talking about branding through the lens of more about like sales and go to market and customers. But you know, from a from a founder's perspective, from a startup perspective, brand is just as important for recruiting purposes. Right To find. You want to. When you go to recruit, you know your initial first engineers and you want the best ones. I mean, first of all, from a branding perspective. You want them to feel like they're joining an innovative company. But then, two, if you have a strong enough brand, you're getting. You're not having to go find those people. They're finding you right, you're finding the people coming to you saying, hey, I heard about your company and I would like to work there because you're a really innovative AI company, and so, yeah, branding, kind of like we've been talking about, affects every part of your company, from sales to recruiting, to culture and all these other things. So I'm going to go a different direction, one that I don't think we haven't talked about yet, but something along these lines of brand makes me think of it is.
Speaker 1:So there was a headline that came out either this past week or the week before where Klarna who's like big payments company they made a big hullabaloo about replacing some of their internal apps with, like homegrown apps they built themselves using AI, and it's unclear to the extent how real this is in Klarna, versus them just saying they did. I think it was the CEO, maybe, who kind of made a big deal of this. But I also wonder if huge name brands like Salesforce, who is a giant brand right and they're known as the CRM If their apps will start to get replaced within companies with people just building homegrown CRMs and people have always built homegrown CRMs, to be honest, like certain companies do build their own stuff, but it's never made sense to do because the bar was so high that building your own homegrown CRM like Salesforce was flexible enough. But if the bar to build your own CRM becomes quite low using AI development tools, then maybe everybody customizes a CRM for their own use case, and so the reason I think about brand in that respect is like there's going to be all these like unbranded, no name apps. They're like popping up internally within companies where people aren't using salesforce. Not all the apps they use are going to be associated with a brand. It's just going to be custom development, custom apps within within a company, and I'll take that one step further.
Speaker 1:I was talking to an investor this week, and where I think that's particularly going to be interesting on the venture world is companies that have huge amount of resources, so like Fortune 500 companies. Right, they are paying huge contracts to the sales forces of the world, to the oracles of the world. I mean these contracts are like $100 million, $500 million. Yeah, if they could just build their own CRM, why are they paying a $500 million contract to Salesforce? And so I think that enterprise but also that's an annual contract.
Speaker 2:And so for them to take a one-time cost to build their own and then do the transition, which obviously would be a massive lift and extremely cumbersome. Salesforce has designed themselves that way on purpose to keep their customers sticky, and Salesforce has designed themselves that way on purpose to keep their customers sticky. But if they are willing to undergo that pain and cost once, then they own it and I think by and large for most enterprise customers they prefer to own their own information.
Speaker 1:Yeah, and I think that'll be really interesting. So I think you'll start to see some pressure on the size of massive enterprise contracts. Some pressure on the size of massive enterprise contracts, because if Citibank is using Salesforce as an example, they could go to them and say, hey, we're paying you $200 million a year, but unless you cut the contract to $100 million a year, we're just going to build our own CRM. Super interesting, yeah, I think that'll be an interesting impact on enterprise software. I think there's a lot interesting impact on enterprise software.
Speaker 2:I think there's a lot to explore here. You're making my wheels turn. This is a position I hadn't considered and I'm just thinking, wow, there's definitely a lot. This almost is a whole episode in and of itself. Such a good point, very, very interesting. And then it's bringing me to two other thoughts just for now, which are one we've talked a lot about AI, reducing development costs and changing the engineering landscape for new companies. We've talked a little bit today and in previous episodes about the AI tools that come into play for marketing and for branding, but what we haven't talked about is, when those are used incorrectly, how that actually really hurts your brand and throws you off course, and so I'd love to hear your thoughts on that too.
Speaker 1:I mean that's funny. You did yeah, you mentioned it before the episode that this is a topic that was interesting to you, and the first thing that came to mind is the Chevy example. I think it was, or whoever it was. They had an AI chatbot that they put on one of their dealerships' websites and then someone made a deal with the chatbot to buy a Chevy truck for like a dollar, so it was like a $75,000 truck and they convinced the chatbot to sell it to them for a dollar. And then they asked it if it was a legally binding contract and the Chevy chatbot said yes, contract. And the chevy chatbot said yes. And so like, yeah, you could do real brand damage by not having the right guardrails in place for your ai chatbots, right?
Speaker 2:and then I think was that intentional, because they thought, oh, we'll take a loss of 70 000 or whatever the dealership or the manufacturer causes, but we'll take a loss of that to bring more eyes to our dealership, because then everybody's going to want to come and chat with the chatbots and then we'll get them through the door.
Speaker 1:I wish I could say they were that smart. I don't think they were in this case. I think they just put out a chatbot as quickly as possible and, yeah, they learned their lesson. Basically, I also think it was like a local dealership, right. So it wasn't like Chevy the brand, wasn't like chevy the brand, it was like a local dealer, but then that gets associated with, like you know, the brand of if chevy or whoever it was. So, yeah, I mean that's. The other thing is, if you don't put the right guardrails in place, you can do some serious, serious damage to your brand because a chatbot could say the wrong thing. You know, a product could produce the wrong image. I mean, a lot of negative repercussions can come to your brand that are generated by AI, and so you have to be really, really, really careful, and I think this is part of the reason why enterprises, cios, are trying to find the right way to thread the needle of adopting AI as quickly as possible, while also doing it in a brand-safe way to not damage the reputation.
Speaker 2:Yeah, I think it's really important because, listen, obviously we both think AI has a lot to offer, but both today, in terms of current state, what it's capable of and even in the future, when it's capable of much, much more. There are still always caveats, and it's important, I think, for people to realize AI can be hugely beneficial, but it can also be equally damaging when used incorrectly. And the way I think about it also is from going back to brand and customer experience and touch points. If you're using AI, especially for anything outbound, and you don't have great controls on that, and all of a sudden your AI system is sending messages to people that sound crazy, that's a brand disaster waiting to happen, huge reputational risk. Sometimes easier is not better. Just being because you can be hands off doesn't mean you should be, and so I just think those are things for people to think about.
Speaker 1:Yeah, I mean, one of the biggest use cases that's being adopted most quickly in addition to outbound, is responding to inbound either inbound sales inquiries or inbound customer success inquiries. And you know customer success, or customer satisfaction, is the most important thing for your brand, right? So if somebody comes to your chatbot on your website let's say you're United Airlines and someone comes in and says, hey, I need to rebook my flight If you have a disaster, if they have a disaster experience with that, then they're going to go on X and they're going to tweet about it and then everyone's going to hear about it and then, all of a sudden, your brand reputation goes downhill very quickly. Yeah, customer support use cases are a really excellent use of AI and also I don't want to call it a risky one, but one you just have to be very careful with.
Speaker 2:I agree and I think that's something we've talked about before and this reiterates it as the human in the loop and I think the human in the loop in all aspects will be the future.
Speaker 1:Yeah and yeah. I think it's like human in a loop and what I would also call appropriate guardrails Guardrails is typically the term that we use in the AI world of like. Can you put the right technical guardrails in place so that, when a response is actually generated by the model, that it actually goes through two, three, four quality checks before it gets to the user?
Speaker 2:And to train your own models that it's okay to say I'm not sure, or for them to not give an immediate answer. They don't have to be people pleasers. Sometimes it's better for them to say give me one minute and then pass it over to a human.
Speaker 1:Yeah, exactly, I totally agree with that, and those are some of the interesting design choices around. Like you know, when do you let your chatbot respond? Versus when do you tell it to say I don't know, versus when do you tell it to pass it off to someone? And like those are like classification tasks of like you're trying to classify the response to say like what's your confidence level of this and if it falls into one bucket, you can respond. If it falls in the other bucket, like this task should go to a human or we should route you somewhere else.
Speaker 2:This has been an amazing conversation. I feel like you and I are on the same page about the importance of brand and I love hearing from your perspective how you think about brand within your company, within the world of AI and then how?
Speaker 1:first AI-created brand, the first brand that maybe is created entirely by an LL, where it comes up with a logo, the branding, what they stand for. I would love to see what that looks like over the next few years. At some point there's going to be a brand that is entirely created by an AI, and I would love to see what that looks like.
Speaker 2:So Claude will tell itself create a new brand.
Speaker 1:Yeah, and then it will run with it. It'll create like the next Nike or something like that. You know.
Speaker 2:Oh, like a meaningful brand.
Speaker 1:Yeah, yeah, I think so.
Speaker 2:Yeah, I think that'll be it. Oh, wow, interesting. I know Nike Now that you said that and we'll talk soon. Yeah, this was fun as always. Thanks, paul.
Speaker 1:All right, bye, bye.
