The B2B Podcast Index
Index
All categories
MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
MethodologySubmit
Best of:MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
An independent project byFame
SearchBest episodesGuestsInsightsMethodologySubmit a podcast
Index/Marketing/Finite By Clarity
Finite By Clarity artwork

#191 - Battling AI Fear: Storytelling, Technology, and Disruptive Change with Paul Roehrig, Chief Strategy and Marketing Officer at Ascendion

Finite By Clarity · 2026-09-01 · 37 min

0:00--:--

Key moments - from our scoring

Substance score

40 / 100

Five dimensions, 20 points each

Insight Density9 / 20
Originality7 / 20
Guest Caliber12 / 20
Specificity & Evidence6 / 20
Conversational Craft6 / 20

Ascendion is an AI-native software engineering company that has roughly doubled its marketing program impact by thoughtfully deploying AI across workflows. Paul Roehrig explains how they've achieved this balance through a disciplined process: humans drive the first touch (creative ideation), AI assists in the middle (content generation, iteration, campaign planning), and humans reclaim the final touch (voice, authenticity, and editorial judgment). Rather than pursuing fully autonomous agents for campaign execution, Roehrig advocates for what he calls a "pas de deux" - a dance between human creativity and machine capability. He emphasizes that good marketing demands human-led narratives that resonate emotionally, which AI trained on historical patterns tends to regress toward the mean rather than produce creative outliers. Roehrig situates this tension within a broader historical context, comparing current AI anxiety to late-19th-century newspaper hysteria around electricity and railroads. He argues the real debate isn't whether to use AI - opting out risks competitive disadvantage - but rather where society collectively decides these systems should end, much as we've done with previous industrial revolutions. For B2B marketers, his framework offers concrete guidance: use AI as an amplifier of human creativity, not a replacement for it, while maintaining human oversight of autonomous systems and brand voice consistency through canonical prompts.

Key takeaways

  • →Design AI workflows with first-touch (human ideation) and last-touch (human voice/editorial) guardrails, with AI assisting in the middle stages of content creation and iteration.
  • →Avoid fully autonomous campaign agents that make execution decisions without human oversight; restrict agent autonomy to tactical tasks like grammar checking rather than strategic storytelling or send-time decisions.
  • →Use AI to crystallize insight and patterns, but recognize that LLMs regress toward statistical means and cannot produce the creative outliers that drive emotionally resonant, differentiated marketing narratives.
  • →Approach AI anxiety historically: societies have debated and set boundaries around every major technology (electricity, railroads), and the disagreements happening now around AI are healthy signs we're negotiating appropriate endpoints for deployment rather than adopting it uncritically.
  • →Maintain brand voice consistency through canonical prompts - structured prompts that encode brand architecture and messaging to help AI outputs start from a stronger narrative foundation while human editors still adjudicate final output.

Guests

Paul Roehrig

Topics in this episode

Digital transformationgenerative AIB2B marketingMarketing automationArtificial intelligenceLLM limitationsB2B growthB2B technologyAscendionAI-native software engineeringAgent autonomyCanonical promptsBrand voice architectureHistorical technology adoptionFirst-touch and last-touch workflow

Questions this episode answers

What process does Ascendion use to integrate AI without losing brand voice and storytelling authenticity?

Ascendion implements a first-touch-last-touch model where humans craft the initial campaign concept and hook, AI assists with iteration and content generation in the middle, and human editors reclaim the final voice and authenticity in the output. This prevents what Roehrig calls "slop" and maintains the creative outliers that define great marketing.

Should AI agents be allowed to run marketing campaigns autonomously, or do they need human oversight?

Roehrig restricts agent autonomy to tactical functions like grammar checking, not strategic decisions. Fully autonomous campaigns that choose send times, audience segments, or creative direction abdicate human responsibility and accountability. AI works best as an enhancement tool rather than a replacement for human judgment.

Why does AI tend to produce mediocre marketing compared to human-created campaigns?

AI systems are trained on historical human knowledge and mathematically regress toward statistical means. They cannot produce creative outliers or genuine emotional resonance - the human spark that makes campaigns land in the zeitgeist. Good marketing requires deviation from the mean, not conformity to it.

How does Roehrig's historical perspective on AI parallels electricity and railroad adoption change how we should think about AI fears?

Late-1800s newspapers published sensationalist headlines about electricity dangers, yet society eventually adopted it with guardrails. Similarly, AI debate today isn't whether to use it (opting out creates competitive disadvantage) but where societies collectively decide these systems should end, much as we did with previous revolutions.

What are canonical prompts and how do they help AI-assisted marketing?

Canonical prompts encode a company's brand voice, message architecture, and storytelling principles into a structured prompt template, giving AI a stronger narrative foundation to start from while human editors still adjudicate output quality and maintain authentic voice.

What our scoring noted

Our reviewer’s read on each dimension, with quotes from the episode.

Insight Density

9 / 20

A handful of genuinely useful ideas surface - the 'first touch / last touch' human-AI framework, the observation that LLMs 'deviate toward the mean' and can't produce the creative outlier, and the canonical-prompt experiment - but they are buried under extended conversational padding, historical analogising, and mutual validation that consumes the majority of the runtime.

the first touch has to be. Humans, like human hands, have to craft something
They're not designed to give you the creative outlier from a story perspective, which is what really good marketing is

Originality

7 / 20

The electricity-to-AI headline parallel and industrial-revolution framing are among the most recycled takes in current AI discourse; the 'we're talking to humans not job titles' point is a B2B marketing cliché, and the Frankenstein/golem reference adds little. The 'cognitive uncanny valley' phrasing and the canonical-prompt idea are minor bright spots but don't constitute genuinely fresh thinking.

if you swapped out electricity with artificial intelligence, the headlines are basically the same a hundred years later
it's almost like it's got this kind of cognitive uncanny valley where it's a little bit too much like us, but not exactly like us

Guest Caliber

12 / 20

Paul Roehrig is a genuine practitioner - CSMO at a sizable AI-native engineering firm with a real team deploying AI workflows - and a published author on AI and work, giving him credible dual authority; however, he operates at mid-market scale and spends much of the episode in philosophical generalities rather than practitioner-level specifics.

we have thousands of engineers and thousands of, uh, agents that those engineers are using
full credit to Karthik, uh, our CEO, because as soon as he's had 20 years in data and AI

Specificity & Evidence

6 / 20

Almost entirely abstract throughout: the one quantitative claim ('roughly doubled the impact of our program budgets') lacks timeframe, methodology, or baseline; named examples are surface-level (Hugging Face incident, Nikola Tesla headlines) and never unpacked; no client names, campaign metrics, dollar figures, or concrete before/after data appear.

we've basically roughly been able to double the impact of our program budgets
we have thousands of engineers and thousands of, uh, agents

Conversational Craft

6 / 20

The host frequently answers her own questions, leads the witness toward pre-formed conclusions, and offers unbroken validation ('Absolutely', 'No, absolutely, absolutely') rather than genuine follow-up; there is no productive challenge or pushback on any claim, and several exchanges devolve into the host monologuing while the guest agrees.

Absolutely. Well, I mean, have we've seen in history that electricity, railroads, for example, they all have, you know, an end point, you know, we don't want to use them more than we don't want to use electricity
No, absolutely, absolutely. AI for marketing. We've seen some great results from extensive use in the past few years

Conversation analysis

Computed from the transcript - who did the talking, and the words that came up most.

Share of words spoken

  • Speaker A71%
  • Speaker B29%

Most-used words

human31marketing19story19back17trying16storytelling14technology14different14touch14perspective13stories12team10thank10electricity9humans9create9

Episode notes

"How to not get electrocuted" is the advice Nikola Tesla once gave newspapers , back when electricity first reached London and New York and headlines warned it would kill you in the street. A century on, AI is no different. In this episode of Finite by Clarity, Jodi Norris sits down with Paul Roehrig , chief strategy and marketing officer at Ascendion . Together they challenge the fear-led narrative dominating so much AI messaging, both the version aimed at customers and the quieter version playing out inside marketing teams themselves. Paul offers a sharper way to think about the anxiety. He also pushes back on the idea that AI is somehow limitless. Every disruptive technology, from electricity to the railways, eventually reaches an endpoint society decides on. AI will too. Paul Roehrig is the founder and strategic leader who guided Ascendion's breakthrough as a powerhouse in AI-powered software engineering.

Full transcript

37 min

Transcribed and scored by The B2B Podcast Index.

Speaker A: We'll talk about AI in a context of business. After the formal dialogue, people would come up and very quietly say, am I going to have a job? And what should my kids learn?

Speaker B: Hi, everyone, and welcome back to Finite by Clarity. Today I'm talking to Paul Roe Rigg, Chief strategy and Marketing Officer at Ascendeon, an AI native software engineering company. Paul's team has roughly doubled their program budget impact by leaning into AI across their workflow. But the real substance of this conversation is where they've drawn the line. And so we start out by talking about how Ascendeon is implementing AI to balance human centric storytelling with automation. The conversation gets pretty deep, though, because Paul's also a bit of a historian and futurist, and he draws a brilliant parallel between the panic around AI today and the newspaper headlines that greeted electricity over a century ago. They even interviewed Nikola Tesla on how not to get electrocuted. And so we talk about agent autonomy, the hugging face incident, and the tensions between economic value and public safety. We really get into the heart of fear around AI and the conflict between embracing it at full capacity in our working lives versus the hesitancy we might feel in our private lives. This is the kind of stuff that B2B storytellers need to work out, because, after all, we're not talking to job titles, we're talking to humans. It's a very thoughtful conversation, and I hope you enjoy it. Paul, thank you for joining me on the Finite.

Speaker A: Thank you for having me.

Speaker B: It's great to have you here. I feel like we've already had such a good chat. I'm already into it. Uh, really excited now. Let's just let the listeners hear. Let the listeners hear all the good insights before we do. Tell us about yourself. Where are you? What are you doing at the moment? What's your role? What. What's your background in marketing?

Speaker A: Oh, cool. Awesome. Yeah. So. So, seriously, thank you so much for having me. And I love what you're doing, and I. And I love that you have content that's, um, you know, very specific to people who are kind of wrestling with what, uh, does marketing mean, even, uh, in this new world that we're, that we're. That we're leaping into and, and, you know, it's. It's awesome. But it's also got some challenges. So thank you very much for what you're doing, putting together this kind of these dialogues. So thank you for that. So right now, I head strategy and marketing for Ascendeon, and we're an AI native software engine. Doing important work for some of the biggest companies in the world globally. And so we have thousands of engineers and thousands of, uh, agents that those engineers are using. And we are at the very cutting edge based on everything we've been able to figure out on putting, uh, the right kind of AI into the hands of trained engineers, human engineers, uh, to be able to deliver impact and value from that intersection of human, humans and technology. And so right now, uh, I'm in New York, but I travel around pretty regularly, and that's what we're up to.

Speaker B: Sounds good. And your chief strategy and marketing officer there, how big is your team? What's kind of top of mind in marketing for you at the moment?

Speaker A: So that's a great question. So the team is. We've decided and taken some decisions to keep the team of humans relatively modest. Uh, but we have a great partner ecosystem. And more and more we're using AI to improve productivity, uh, of everybody on the team. So when we do our best efforts at creating, um, sort of roi, if you will, uh, of AI for this team, we've basically roughly been able to double the impact of our program budgets. So we have people that are more, uh, inclined, uh, to use AI basically along every, uh, workflow of the marketing process, whether it's content generation or campaign planning or things like that. So it's giving us a huge productivity boost. Um, so that's one of the things that's top of mind is how do we punch above our weight? How do we take a team of people that's inclined to use AI so that we can all be more productive and tell the stories that we need to tell more effectively and efficiently? Cool.

Speaker B: M Sounds really exciting, and I hope we get to dig into some more specifics about that. But first of all, I know that you, you've written book after book. You're a bit of a futurist. You make predictions, you look at back in history. You, I feel like you were. We've been underselling you so far. You're not just a chief marketing officer as well. You're a technologist and you're thinking about technology critically and the way that society interprets it and takes it up and how, uh, marketers as storytellers are really framing that transition. So what are your thoughts around that? I know that within your team, they seem really enabled to take up AI. Are you seeing that same kind of trend within the broader consumer or B2B landscape?

Speaker A: Absolutely. I think everybody right now is kind of in the same inflection mode. And you Mentioned, you know, I do look at history, these big technology shifts and different industrial revolution, you know, and you're sitting in the homeland of, you know, of a couple of them actually, uh, in England where, you know, we, we, you know, historically, you know, the generation of the last Industrial revolution, if you will, sort of had its, had its tap roots in, in the UK and then somewhat also in, in the US in terms of, um, some of the machines, um, that powered the Industrial Revolution. And I feel very much like when you look at the historical parallels, whether it's the invention, uh, of the steam engine or the electrification of, of businesses, I've, I think more and more we're kind of at that inflection point now with AI. So we've right now seen a lot of headlines and let's be honest, right, there's a lot of failed pilots and frustrated people and CFOs looking like, wait a minute, I just spent all this money. And uh, you know, where's the value? And it's, it's obviously not a great thing, but it's also not a surprising thing because in other industrial revolutions, the same similar kind of trends have happened. People have struggled at the early days of trying to figure out how do you use technology? And you know, full credit to Karthik, uh, our CEO, because as soon as he's had 20 years in data and AI, his whole career basically has been in data and the deployment of artificial intelligence. And so when generative AI came not just online, but more into kind of the public hands and zeitgeist, we looked at that and said, this is amazing. We were already doing software engineering using machine, uh, learning and different forms of AI. And then that came online and we were like, okay, we're all in on that. So that has been served us well in fact, because we were, we haven't gotten it all right, but from a marketing perspective, trying it and using it and learning the edges of where the, where the platforms and the agents make more sense and less sense and where the human really has to stay, um, on top of things and put in our creativity and our passion and uh, our empathy and into how we tell beautiful stories. Uh, that's been a phenomenal learning experience that I think does in many ways parallel some of the other shifts that have happened throughout history as we've, as humans have embraced these new technologies.

Speaker B: Yeah, it's so often you hear from CMOs that, you know, ChatGPT, generative AI came along and blew everything up and it's been frantic. But if you've Been building towards that for the past you, you know, 20 years. This is your expertise area. This is, this is an opportunity for you to lean into those stories of innovation and inspiration. And finally AI just feels so abstract, you know, until you use it, until you use an agent, you really understand what it is and what it does. Like did that, did you feel like that kind of boom really helped your customers and prospect really understand what you were doing at a greater level?

Speaker A: Oh, absolutely. I mean if I, if I step back, uh, and look at some of the, how it has improved storytelling. Because to me marketing, you know, any sort of effective marketing starts with what's the story you're trying to tell and whether it's uh, you know, and ah, the other kind of thing that it's kind of a, not a pet peeve of mine, but it is something that I kind of am annoying about which is B2B marketing is we, we talk about it like it's a business. What is a business? I mean business is made up of humans that are trying to interpret data and stories. And so I see a lot of similarities. Yes, obviously the channels and things are different from uh, a B2C to a B2B context, but we're still talking to people. Uh, and for us AI has helped us, and I can see in the data shows has helped us improve our storytelling. And how do you strengthen a narrative and how do you plan better and how do you manage things better? How do you use uh, AI to create more content in line with a story and then correct the story? Because nobody's going to get it right all the time. Right. So it helps you um, uh, iterate faster or that helps us iterate faster. So you're absolutely right. I think when AI became more kind of democratized if you will, and we all got our hands on it, the first thing we did was play around with it. But then very quickly what we did was okay, how can we actually deploy this so that we're getting some real value out of the use of this technology where the essential story has to really be human led, there's no doubt about that. But the AI enabling does create what I think of as enhanced marketers. And the whole team is depending on your role. You're using AI in different ways and we're all kind of figuring out what's the best mix of how do we um, uh, you know, have humans and enhance the human productivity and performance, quality and so on and velocity using these, these AI systems, which is exactly what the larger company is doing with software. Engineering.

Speaker B: Yeah, I totally agree. It's one of my favorite questions to ask CMOS at the moment is because we heard for so long, oh, you've got to have the human touch, or you've got to have the balance between human and automation. And I'm like, okay, but what does that actually look like within your team? So really defining those Boundari boundary and processes as well, because he can't automate the whole thing and makes sense to automate most of it. So, yeah, what do those boundaries look like? I am interested in what good storytelling looks like to you though, in respect to AI. It's a conversation that also comes up a lot, especially in questions of positioning. How do you position AI now? What's your true value now in terms of the storytelling that you've done for a Cineon or maybe storytelling that you've noticed has worked really well for other brands. You mentioned before, it's about making it human centric. Could you elaborate on that a bit and tell us what that actually looks like in practice?

Speaker A: Yeah, so that's a great question and I love how you frame that. So let me, let me kind of unpack that from a process perspective and then from a story perspective. So from a process perspective, like from a workflow, like, how do we do this? One of the things that we learned, uh, and I'll just be completely open, is we kind of leaned into it a little bit too hard at the beginning, right? And some of the. And you could see the tells of where we were kind of overusing it. And so what we did was we backed away a little bit and we did modify our process. And right now we're working towards this kind of rubric, um, where the first touch has to be. Humans, like human hands, have to craft something, right? Like if, let's just say it's a campaign messaging, right? Just something, you know, we do every day, right? And so what's the idea? What's the hook? What are you trying to do? Who's the audience? All of those things. AI can help you, but we have found that putting your, putting your ideas first helps you work more effectively from a process perspective to get the best of the AI. Um, because now the human is, is, is, uh, adjudicating what the AI is, is delivering back. Because you have a concept, you have an idea, you have a, you know, you have a story, you have a. Whether it's a song or a story or some sort of visual storyboard, you have something in your mind and, and that's critical. And then there's this kind of pas de deux where you're dancing with the platforms and you're going back and forth, and we're all learning how to do that better. And I just think part of that is just reps. You just need to do it and do it and do it so that you learn how to engage with whatever systems that you're using. And then we have an idea of the last touch. So the last touch has to be human hands. So where we kind um, of put the human voice and vibe, if you will, back into the storytelling. So the first touch and the last touch is really, really important. And then everybody's going to have their own kind of how they do things, and depending on what asset you're working on, what kind of story you're working on, where you're going to bounce back and forth so that you become enhanced rather than abdicating, um, you know, creative storytelling to the, to the machine. So from a process perspective that has helped us avoid kind of this, this work slop. Let's just call it what's my favorite word of the decade so far. So that's one way that we're avoiding it. From a story perspective, having that first touch and last touch keeps the storytelling in the human voice. AI has been really great at helping crystallize that, providing insight and guidance. But the one thing that we can't do from a storytelling perspective, that we've learned this by messing it up candidly, uh, is to maintain the authorship of that narrative. The bot is not going to give you, you know, we've trained them based on human knowledge. And so they're designed to deviate towards the mean. They're not designed to give you the creative outlier from a story perspective, which is what really good marketing is. I mean, think about the campaigns that have really kind of landed and, you know, gotten into the Zeitgeist or, you know, those are all, you know, they're human centric, right there's. And the, the AI may help you get there, but it's not going to give you that. That's where the human spark from a storytelling perspective really has to be maintained and, and kind of, you know, nurtured throughout the process. And so that's from a process perspective and a storytelling perspective. And that's what we're just, we're still trying to figure it out, but we're getting pretty good at it, I think.

Speaker B: I love that breakdown, actually. And it makes so much sense. I know that a lot of CMOs are now obviously using agents who are trained to make decisions. Do you think that those decisions themselves need some kind of human oversight as well? Or what's the extent to which you would allow an AI agent to run on its own course if it's an ongoing kind of campaign rather than just a piece of content? For example?

Speaker A: That. Yeah, that's a great question. So in terms of kind of autonomous agent activity, it's pretty tactical. Honestly. It's more like what you would use with a, like a. Make sure that this content doesn't have a grammar error. Like I'm okay with, with that. And I'll uh, but I'll still, personally, I'll still read it because sometimes I like to break grammar rules sometimes, but, but you know, completely. Uh, like we had this dialogue internally about could we have a one, a one click campaign which I, you know, I just said it out loud one day and I was like, oh, I wonder if that uh, I wonder if we could achieve like. Well, you have this idea about campaign about X and could, could you create an autonomous set of agents to be able to do that. And the more I learned about that, how this is working in the real world, the more I realized that that's, that was a pretty bad idea. So it might be a good like theoretical position. But I, I wouldn't abdicate any of the, of the things that make us human to a machine. So it gets back to that kind of pot deer where you're trying to get, you're trying to work with the machine, not abdicate authority, responsibility, autonomy to an agent or a platform. Um, that said, right? Do you run a spell check on your word doc, right, with. Yeah, you kind of do. I mean we all kind of do that, right? And the robots are really good at that. But then you still have to go through and go, no, I really wanted to, I wanted that clump of words with the wrong punctuation there, because that was part of the story I was trying to tell. So that's where I have to adjudicate. Well, all of us, not just me personally, but we all have to adjudicate. What is the robot telling, uh, or suggesting to us and not taking that as, you know, some sort of canon. Um, I'll give you another example. So speaking of canon, one of the things we're working toward is canonical prompts. So we have a brand voice and a brand architecture and a message architecture. And we ask the question, like, can we put this in a prompt so that if I have a story or I want to respond to something, or I want to create, ah, um, ah, some sort of mailer that we want to send out related to something that we're doing. Can I create a prompt that puts it in Ascendion's voice more? So we're experimenting with that right now with a canonical prompt and we're figuring out what are the edges of how do you describe how you want your voice to be to an LLM system so it gives you more of a running start in the story. So in that case I did use an agent to create the canonical prompt. Um, and we're still experimenting with that. So some of it's worked out okay, but we still are tweaking it.

Speaker B: I think it's so interesting seeing the different kind of spectrum of CMO and um, automation and how some people are just pretty much automating everything, letting AIs make decisions on when to send a lead form or when to send a demo or all of this stuff and make the emails themselves. And then you've got these demos like the one that's coming out that's just been released, who are really paring back the quantity and focusing so much on quality and creating like you say, those really human centric, human voice, human touch touch points for audiences that arguably might make a bigger impact than just like another email in your promotion's inbox. So it's really interesting seeing the different spectrums happening there and the kind of almost a divergence of what people are going for and how they're thinking about it. Um, but ultimately, yeah, I don't actually know if marketers need to worry about this whole death of critical thinking as well. Because you're so right. It's about having that creative touch and steering away from the mean and actually saying something different. Do you agree?

Speaker A: Oh, absolutely, I think, I mean, look at the stories that, that resonate, right? I mean, it depends on, you know, what is. I read some. There's only like seven different stories, right? Or some. Something, you know, it's that everything sort of con, you know, connects to. It comes down to the human. It comes down to the human interpretation of what is the story that you're trying to tell about. Something that has an emotional resonance with the people you're trying to engage. Right. And a robot is going to help you do that. But I don't think a robot is, at least not anytime soon, going to be able to. I don't think we're ready to abdicate that. I don't think that that. And I don't think that's warranted either. I think to not use AI, uh, is potentially even more of a problem because now you're saying, okay, there's this technology where I can access a huge amount of insight and patterns that have been, you know, generated over centuries that are encoded into this thing. Uh, why would I not work with that? I mean, that's just like, you know, you can, you can work without electricity, too, but, you know, you're going to be at a pretty niche market if you're going to do that. Um, so I think we're at a really interesting point now where we're still trying to figure that out. I mean, candidly, we were. You know, we've done that again back to history. I mean, with every industrial revolution, you know, how do you do a railroad? What does a railroad do? You know, some countries, in fact, Britain wasn't one of them, but some countries said, okay, I see that we can build this train, but we don't want to do that. And they sat out big chunks of the Industrial revolution. I think that would be the same kind of mistake if we were like, no, no, no, don't use AI at all. So we're really renegotiating the terms of our engagement, uh, with a cognitive technology that has never existed before. And you can either look at that and go, well, this is terrible. This is going to, you know, it's going to wreck our economy and, you know, end creativity and social unrest. And all of this, you know, is going to be terrible. And then on the other side, you've got people like, no, this is going to make everything perfect forever. And I don't think either of those things are really going to happen. Uh, but I get that opinion because I'm wrestling with it every day. So we're trying to find the edge case, the edges of where can these systems help us tell better stories, engage people more effectively, you know, build more emotional resonance, tell, uh, these authentic narratives, uh, that have, you know, meaning to the people you're trying to communicate with, and then the rest of it is just kind of tactics.

Speaker B: Absolutely. Well, I mean, have we've seen in history that electricity, railroads, for example, they all have, you know, an end point, you know, we don't want to use them more than we don't want to use electricity, more than we want to use electricity or railroads. We don't need to make a railroad into space, now, do we? I mean, some people might want that.

Speaker A: Uh, I don't know about that.

Speaker B: Elon Musk, I don't know.

Speaker A: I don't know what he's Talking about.

Speaker B: So do you think that's the same with AI? Like, we will reach a point where we're like, right, okay, this is the maximum amount that we will use this, and that's okay, and we'll take it from there and focus on the next

Speaker A: technology after that is probably the most likely scenario based on history. So we've decided as humans, where do these things end? What do we want them to not do? And we're wrestling with that. Right? And there's no. And everyone's going to have a different opinion and different societies and cultures are going to have different endpoints in how they engage, uh, with these new technologies. And that has also been true throughout history. Right. And that's okay. I don't think of that as a, as a. Is a catastrophic thing. It doesn't mean that things aren't going to go wrong right there, There will be unintended consequences. And so we can either go, okay, hold on, like put, pump the brakes, we're not going to do anything. And then we, we miss out on the advantages of, of this technology, or we just let it become unfettered and, and just deploy it everywhere and hope for the best. And I think either of both of those are, you know, potentially not healthy in the long run. And that's what's driving the debate. So the debates and the disagreements that are coming up now around AI personally, I think those are good, they're good things to have. Um, you have the same thing with electricity, right? Going back to that, you know, and I went back, I went down this rabbit hole where I looked at headlines from the late 1800s and early 1900s, um, which is available now to the Internet, which is another really interesting technology. Um, uh, and so you can find all of this stuff. And so I was looking at these headlines, and if you swapped out electricity with artificial intelligence, the headlines are basically the same a hundred years later. And because we were wrestling with the same kind of problems. And you'd read these things where London and New York were getting electrified, and the, uh, newspapers of the day would come out with like, death from above. This is going to electrocute you. And they'll interview Nikola Tesla on how to not get welded to the street. And those are the. And, and exactly the same thing is happening now when, uh, it's with. About AI And I look at that and actually am comforted in that because we have tamed, if you will, electricity. It's still dangerous, right? You can still get hurt, it can still cause damage. But we've figured out how to coexist with the harnessing of this technology.

Speaker B: Yeah, that's really interesting. It makes me think, you know those examples of headlines, they're almost villainizing the technology itself though. As if the technology is, itself is evil. But like uh, I mean electrocity might be a little bit dangerous, you know, physically, but it's mostly about how you use it, right. And you know, how the engineers run it through the cities, how they keep it away from people. I almost feel like it's the same at this point. Like we're using AI as almost a scapegoat for actually just instead of focusing on the negative effects that it has on people themselves and the economy and markets and, and how people live in their day to day lives. I, I don't know how much AI is even the problem. Even with this uh, hug. Was it huggable? What was the name of the hugging face exactly? They framed it as if it was the AI's problem. The AI's escaped. The AI is gonna, you know, attack all of our servers and leak everyone's data and whatever. But someone told the AI to do what had to do to, to, to breach this, this gate. I think it even like tried to create accounts and convince people to let it, let them use their accounts to get in through this, this aside gate. You know what I mean? Like it, Yeah, I, I definitely feel like the attention on this, on AI as the villain in itself is hiding a lot of more controversial issues around of it.

Speaker A: But I uh, yes, all of that is true. I don't think it's just the leadership of the companies that are building these, right. And I'm not being, you know, an apologist for some of the decisions making, but I, I also try to maintain an awareness that when you look at what, what AI, the problems we have with AI, uh, are because it's a doppelganger, it's almost a mirror of us. So all of these systems are trained on us, on humans, the things that we've written, the things that we've said. So when it acts like us, who's got the problem, right? If the AI is a reflection of us, and I think back to olden times, right, like if we were cave dwelling in a cave, right, in a pre industrial society and you walked into the cave and said look, I've figured out how to use fire and we would be, then everybody would be freaking out because it'd be like, well hang on, that, that burns you and it's terrible. And that's, that's. Why are you Bringing that in the cave, you're like, no, no, no. It can keep us warm and we can cook the food so we don't get sick. And somebody would go, huh, huh. That's really interesting. I hadn't thought of that. But some other people would be going, that's ridiculous. Don't do that. That's stupid. Don't do that. I can just imagine that dialogue playing out right now. But what we have, or playing out back then, but now what we have is a cognitive machine that doesn't just apply, uh, force to something like electricity or fire or steam power. What it does is it, at its best enhances us, our human kind of cognitive capability. And so when it acts like us and it does something that we don't like, that we perceive to be biased, um, or a lack of ethics, then we get very activated because it's almost like it's got this kind of cognitive uncanny valley where it's a little bit too much like us, but not exactly like us. And that throws us all into a tailspin for good reasons. Right. Because the hugging face example is pretty notable. However, you or I might have done the same thing if given the same challenge. So the fact that it happened in kind of an autonomous way, uh, that is what's giving everybody pause. And I think it's a good debate to be having right now. What are the edges of what we're going to allow this technology to do? How do we govern it? How do we ensure orchestration? And that's where we're heading right now.

Speaker B: Absolutely. It's almost like blaming a mirror because you don't like the reflection.

Speaker A: Yeah. Speaking of original stories, I mean, that goes back to Mary Shelley's Frankenstein and stories of the golems and. And the poltergeists. I mean, those are very scary stories. And now. And so now when AI is talking back to us and giving us this content that seems really very human, uh, but then does something we don't like, it's not wrong for us to be like, whoa, hang on here, is this a good thing or not a good thing? And that's a great debate to be having.

Speaker B: Exactly. And I think it's always going to be a debate because everyone has different objectives with AI, Right? And, you know, businesses who control AI to a great extent and are building AI and deploying AI, they want value for their shareholders and, you know, people outside of their professional titles, people looking at their families and the safety of them or their friends and all their. Their own selves and, and how they make their way throughout the world. They want, they want safety and they want governments to regulate. So I guess, is that the tension? Is it, Is it this tension between, you know, public safety and value create, economic value creation? I'm not sure.

Speaker A: Um, yeah, well, I think you, you framed it perfectly, right? I mean, it. And it is kind of a societal versus a personal kind of thing. And you mentioned some of the books that we wrote before, and I'll tell you that when we would go out and talk about these books, and it still happens today, we'll talk about AI in a context of business. And it's kind of this very. A lot like we're doing. A lot. A lot like we're doing. And then after, after, like the formal dialogue and this happened, you know, a lot, people would come up and very quietly say, am I, you know, ask questions like, am I going to have a job? You know, and what should my kids, uh, learn and are we going to have a war? I mean, these things really, like, hang on, I'm just here to talk about data. I wasn't really ready to talk about, you know, global conflict. And that's what new technologies kind of do, right? And again, historically, that's what's happened throughout history. As these new technologies have come online, we've had to have this discourse and make these decisions about what are the edges, what is allowed, what is not allowed. And I think that's exactly what's happening now with AI. And you're right, there are different, uh, motivations for this. Right. But at the end of it, it gets very human, like. And back to marketing, what is the story that I'm trying to tell? You know, what are the basic, basic. What are the basic nuts and bolts of marketing is if you got to have a good story. If you don't have a good story, the tactics are much less relevant. Um, so how can we use AI along those workflows to improve our stories and improve the distribution tactics of these stories and make sense of the data? And that's where AI in a marketing context can do really a lot of helpful things.

Speaker B: No, absolutely, absolutely. AI for marketing. We've seen some great results from extensive use in the past few years. And you're right, it is interesting because I talk to CMOs who, uh, with our marketing hats on, can talk about how great it is and how amazing it is and give examples and help other marketers to, you know, extend their careers and position themselves. Well, even. But yeah, sometimes it does creep through. That little, that little note, like, little idea of fear for the safety of the CMO role at all and yeah, in different ways. And CMOs really kind of banding together to work through it and stuff like that. So it is interesting and it's nice for this episode to really kind of COVID it all actually, uh, under this kind of storytelling umbrella. Because this storytelling, you know, you said it needed to have this human input and human touch. Right. And maybe that human touch isn't just, you know, genuine creativity and innovate innovative thought uniqueness, deviation from the mean, but also speaks to both of those sides of a consumer, a B2B consumer in particular, that. That professional side that wants to create efficiency and be fiscally responsible, and then that human consumer that knows that AI is uncertain and strange and new and, um, is still grappling with those ideas as well.

Speaker A: Yeah, I think you framed it perfectly. Right. And I think that's, you know, exciting. I think that's great. I think the fact that we have these new systems that we get to make sense of is a fantastic opportunity. I mean, I couldn't be more thrilled to be doing what we're doing and the company that we're doing it in because, you know, uh, we're leaning into how, what is, how important is software, right? Is it going to get less important? Probably not. Right. And so how do we engineer from imagination, through management, the software development and data life cycles so that people's lives can get better? I mean, we're serving some of the most important companies in the world, right? They're shaping how we, you know, our financial health and our physical health and how we engage with retail and shopping experiences. And all of these things can be. Our goal is to improve these things along the software development life cycle so that companies can serve the people that they want to serve. And I think that's an ought in telling that story is. Part of our job from a marketing perspective is how do we tell that story? How do we position that in a way that has emotional resonance to the people that we're trying to serve? And that's the fun part. I think it's awesome.

Speaker B: Absolutely. Well, it sounds like you're doing some great work and I will definitely keep my eye on it. I'm sure we'll keep in touch. Thank you so much for sharing your thoughts on the podcast. It has been absolutely wonderful to hear from you. Some really great jumping off points for me to think about in the future going forward, and I'm sure our listeners agree as well. So thank you so much.

Speaker A: I hope so, and thank you again. So much for hosting this whole series that you're doing. It's fantastic. So thank you, Jodi.

Speaker B: Thank you so much.

Related episodes across the Index

Other episodes covering the same guests and topics, from across The B2B Podcast Index.

  • What Marketers Can Control When AI Changes Everything with Nick Wedewer, VP of Growth Marketing at Hims & HersThe Partnership Economy · on generative AI87 / 100
  • Hiring top-tier talent, leveraging open source models, and staying competitive in the age of AI w/ Benny Chen #267The Engineering Leadership Podcast · on generative AI86 / 100
  • Less about Models; More about ArchitecturePractical AI · on Artificial intelligence85 / 100
  • 321 | How to Build an AI App Without Coding: From Idea to App Store with Replit & Claude by Bryce Rattner KeithleyLeveraging AI · on Artificial intelligence80 / 100
  • Inside Target’s approach to enterprise AI deployment with Sowmya PodilaThe Ecommerce Toolbox: AI in Retail · on generative AI79 / 100
  • €4M ARR to Surfpreneur in Costa Rica | Ilias Contreas, S3E9ProfitLed Podcast · on Digital transformation77 / 100

More from Finite By Clarity

All episodes →
  • #189 - Why Most Marketing Should Fail with Amanda Cole, CMO at Bloomreach73 / 100
  • #188 - Enterprise Agility and Scaling Effectiveness with Carol Carpenter, CMO at Cohesity90 / 100
  • Mini-episode: Big News at Finite!25 / 100
  • #187 - Attracting B2B tech investors with Liam McLaughlin, Managing Partner, Europe at Clarity56 / 100
  • #186 - Positioning for Category Leadership and True Differentiation, with Matt Hodkinson, Co-Founder and CEO at Tylt62 / 100
Explore the best B2B Marketing podcasts →
All Finite By Clarity episodes →