What’s the BUZZ? — AI in Business

Generative AI In The Enterprise — Trends & Predictions (Guest: Daniel Faggella)

Andreas Welsch Season 3 Episode 1

In this episode, Daniel Faggella (Founder, Emerj Research) and Andreas Welsch discuss the big trends and predictions for Generative AI in the enterprise. Daniel shares his insights on Generative AI and provides valuable advice for listeners looking to implement AI in business this year

Key topics:
- Anticipate key trends for Generative AI in 2024
- Determine focus areas for AI leaders to prioritize AI this year
- Drive adoption of AI and Generative AI in your business
- Outlook on Generative AI over the next 3-5 years

Listen to the full episode to hear how you can:
- Understand why established vendors will lead the first wave of Generative AI
- Benefit from an increased level of AI fluency within the C-Suite for AI adoption
- Assess risk appetite vs. risk aversion within the enterprise
- Expect a significant impact of LLMs on work and the economy


Watch this episode on YouTube:
https://youtu.be/nbDSvfBClqw

Questions or suggestions? Send me a Text Message.

Support the show

***********
Disclaimer: Views are the participants’ own and do not represent those of any participant’s past, present, or future employers. Participation in this event is independent of any potential business relationship (past, present, or future) between the participants or between their employers.


Level up your AI Leadership game with the AI Leadership Handbook:
https://www.aileadershiphandbook.com

More details:
https://www.intelligence-briefing.com
All episodes:
https://www.intelligence-briefing.com/podcast
Get a weekly thought-provoking post in your inbox:
https://www.intelligence-briefing.com/newsletter

Andreas Welsch:

Today we'll talk about predictions and trends that will shape AI this year and in the years to come. And who better to talk about it than someone who's got excellent insights into that. Dan Faggella. Hey Dan, thanks so much for joining.

Daniel Faggella:

Good to catch up with you again, brother. Glad to be here.

Andreas Welsch:

Wonderful. Hey, why don't you tell our audience a little bit about yourself, who you are, and what you do?

Daniel Faggella:

I'm Dan Faggella. I run Emerj Artificial Intelligence Research. We're a market research and publishing firm focused on the ROI of AI. So we've got the biggest podcast in sort of B2B AI called the"AI and Business" Podcast, like a million and a half listeners every year. Andreas was a guest with us a good number of years back there. So good to be able to meet you on that. And our focus is less on helping the folks that write the code, although I think that's really important. Our focus is more on what is the P&L impact and the workflow impact within the Fortune 500? So whether it's talking to the CIO of Goldman Sachs or the head of AI at a Sanofi or something like that. Direct primary interviews with those sources is what we publish. And then we help big vendor companies that want to reach the Fortune 500 sell into those big orgs. So we're always touching the buyers and the sellers at a high level. And that's my full time job.

Andreas Welsch:

I'm always amazed at the guests you bring on onto your podcast. Definitely check out Dan's podcast. I think it's, definitely one of the best ones around. And I really enjoy reading your perspective online. I know you post a lot. Yeah. It resonates deeply with me. So I'm having a bit of a fanboy moment here. I must tell you.

Daniel Faggella:

I've been following this stuff since you first started publishing and happy to be here.

Andreas Welsch:

So folks, for those of you in the audience, if you're just joining the stream, drop a comment in the chat where you're joining us from. I'm always curious to see how global our audience is. Dan what do you say, should we play a little game to kick things off?

Daniel Faggella:

Absolutely. Let's go for it. And I told you, let's keep it a surprise. So I'm here to be surprised.

Andreas Welsch:

All right. So this game is called In Your Own Words. And when I hit the buzzer, the wheels will start spinning. When they stop, you see a sentence, building up the pressure and the tension.

Daniel Faggella:

All right, let's do it.

Andreas Welsch:

So I'd like you to answer with the first thing that comes to mind and why. In Your Own Words. To make it a little more interesting, you only have 60 seconds for your answer. those of you in the audience watching this live, feel free to pop your answer in the chat as well. I'm curious what you come up with, too. Dan, are you ready for What's the BUZZ?

Daniel Faggella:

I'm ready, brother. Let's do it.

Andreas Welsch:

Dan, let's go. If AI were a bird, what would it be? 60 seconds on the clock. Go.

Daniel Faggella:

Oh, good gracious. By the end of 60 seconds, I'll get this for you. I promise. It would have to be a very versatile bird for sure, because it's certainly not occupying only one part of the ecosystem. I'm torn between two. On the one hand, I feel like it could be an eagle or a hawk. Really defining the skies where it lives. Many other species are responding to it because it wields a lot of influence and power. At the same time, I feel like it's like a rock dove a pigeon because the more humans get together and the more human activity there are in buildings and whatnot, the more data, if you will, spills out of humanity. The more they can cluster, they can live next to us, live around us, live, live, off of the ecosystem that we construct. So if I was going to say, a hawk or a rock dove, I hate to say I probably have to pick the pigeon, but not because it's not powerful, but because it's so tightly entwined with man. So that's my final answer. I'll end with that one.

Andreas Welsch:

That's great. We've had a Phoenix before. We've had a Condor. We've had an Eagle. I love how you explained that it's so closely related to us.

Daniel Faggella:

And where humans congregate. That's where they, all come together. They flourish. Yeah.

Andreas Welsch:

Yeah that's awesome. Dan, why don't we jump into the questions that we talked about before? And again, thanks for answering this one on the spot. So look, I think we've all seen that Generative AI hype and shoot it up so quickly the hype cycle last year. I'm wondering, what do you expect we'll see this year, right? If anybody has a crystal ball or can make a good guess, I figured that'd be you, talking to all these leaders across so many industries and so on. What are you seeing?

Daniel Faggella:

It is really tough of course to say what breakthrough where, but there are directions that feel like they're unlikely to reverse. So when we study things that Emerj, when we're publishing, when we're connecting the dots between Fortune 20 banks and the people that sell to them, or whatever dots we're connecting with direct interviews. Sometimes there's individual use cases we're excited about. Sometimes there's just general shifts. We refer to them as inevitable. And, it seems like there's a couple things that I would expect and we can unpack whatever is relevant for you. There's a couple of things I would expect are going to go down. So we talk about the enterprise first. I can tell you with robust and amazing high levels of confidence that the gen AI ecosystem, startup wise, venture backed wise, is really not enterprise ready right now. I can tell you that because when they get enterprise ready, they're in the inbox. And or they're already talking to the people we're talking to because we talked to the people that spend the dollars in sort of the Fortune 500. And they're really not there yet. So the first wave of cash is going to the services firms, that already have embedded relationships. And then there's a whole bunch of other folks that are somewhere in the middle, like software firms that have big bases and they have services that sit on top of them. We think that those relationships with people that kind of can flexibly build for the buyer. That's going to capture round one of value because the exciting venture backed stuff is a little bit too abstract. It's not close enough to workflows. They don't even have sales forces yet. And a lot of it's getting built for consumer. So I think our first wave is going to be with incumbents that really rake in the cash around Gen AI. I don't know if this is going to be the year where we see a million blossoms bloom for Gen AI capabilities, but I do expect it to be a much more fruitful blooming than we saw with, let's say, the chatbot wave five years ago, or other kind of hype waves we've gone through. Because I think we have a more AI fluent enterprise crowd, and I also think we just have a more capable technology. I gotta be frank with you. I think we're itching towards the, we're itching towards the big stuff here, brother. Like the stuff that really changes sort of the human condition. And so I think we're going to see some movement. But in the enterprise, those are things that I would expect to happen in the next year. And I've never laid out a crystal ball, but those are things I expect. On the Gen AI in life side, I expect there will be more and more consensus around the fact that we are going in. In other words, I think that it's going to be extremely compelling, to conjure a 30 minute video, even if by the end of next year, it's only a 10 minute one, by the way, or a four minute one based off of a verbal prompt that may be unbearably attractive from a learning perspective, right? If you want to learn a skill. It may be unbearably attractive from an entertainment perspective. You better believe every single social platform already gets this right. Eventually it'll be conjured to capture your specific attention as an individual user. And then there's the kind of the dating side of things, which I won't get into vastly more detail, but you can let your imagination run there on that set of mental circuits. I think that it will become somewhat self evident that actually particularly Gen Z, but humanity at large, we have to wrestle them into the real world or kind of accept that a lot of things are going into the virtual world and we'll just be more fulfilling and richer and better there from an experience standpoint. I don't mean that as a dystopia. I mean that as a real force that we have to contend with. And I think that humanity writ large is going to have to say, by golly the human condition is different. How are we going to make sense of it?

Andreas Welsch:

Thanks for sharing that. I think it's a great balancing on one hand what you foresee will happen in the enterprise or what you already know is happening in the enterprise in this wave one and also what's the impact going forward on society and people at large. Thanks for sharing that. So also for, those of you in the audience, if you do have a question for Dan, feel free to put it in the chat and we'll take a look in a couple minutes. Now you already touched on the enterprise quite a bit. And I'm curious there, what do you think enterprise leaders need to focus on a lot more this year than they have been doing last year or over the past couple of years? What really changes for them? What do they need to be aware of?

Daniel Faggella:

Yeah I would say the fundamentals are the fundamentals. Skills, resources, culture. We have an article, you just type in E M E R J critical capabilities. So these three main clusters of these things are going to have to get built up if you want to get value out of AI. The foundations really still have to be there. So I would say it's not throw out the playbooks everybody. It's like now if you've learned some hard lessons about cross-functional teams coming together, about investing in talent, about allowing for R& D on the culture side, around getting your data house in order with infrastructure. Like probably don't throw that stuff out actually. But I would say a continued, robust investment in critical capabilities. That's going to continue. In terms of what's new, though, good question. There's an element of executive AI fluency. There's an element of that which is familiarity with the range of use cases. So knowing what can this stuff even do. And, executives having a grasp of that, it's very important for them to have a grasp of that to know the two elements of fluency. What conceptually can AI do and not do? And then also, where does that fit into strategy? You really can't fit AI into strategy if you don't know what it does conceptually. You don't know specifically what it can do in the world. With Gen AI, There's lot more potential to be able to get the simple lowdown of what it can do, right? Explaining the really dark nuance of identity verification for like anti money laundering is possible. And, there's articles and got interviews and all kinds of stuff, but it's very different than here's a bot spinning up a two paragraph summary of a one hour long interview with insights specifically focused on accountants. That's a very snappy, concise way of staying abreast. So I would say whether whether you use Twitter or LinkedIn or what have you, whether you get yourself on some good newsletters being able to stay abreast of what is conceptually possible I think is more accessible to the imagination than ever. Number one, there's more news sources. When I started doing the podcast 10 years ago, not that many people have cared about AI, honestly. But now there's a lot more people. And secondly it's a lot more visually appealing and succinct in terms of some of the areas of Gen AI, not all of it, but I think some of it. So those would be two things that come to my mind that I think leaders should. Stay pretty well invested in.

Andreas Welsch:

Now, I'm curious, building on that, I think last year we've obviously seen a lot of excitement around Generative AI with all the loom and gloom and doom and everything in between. But also lots of people and organizations trying this out. On one hand, like you said, as a consumer, I can go into ChatGPT, I can go to MidJourney. You name it. And, I can fiddle around with these things and I get pretty good results quickly. So I can touch the technology. But then also I think this year will likely be the year where the rubber meets the road. And, you cannot continue just doing POCs and proving things out. At some point you need to move it into production, into a real environment. I'm curious. What are your thoughts there? What are you seeing? What would be critical?

Daniel Faggella:

Absolutely. So a bunch of things coming together. We are seeing more willingness to look for kind of partners on the strategy side of AI, as opposed to just point solutions for specific band aid problems, which is really what AI was for what the last eight years. So since the first company started raising money and had AI on their homepage, the first round of them, for the first three years, they were all like, we're going to revolutionize insurance. We're going to revolutionize the call center. And what they learned was, number one, you went to Stanford for four years. You don't actually understand insurance. And so you really need to understand the problem. Number two, the enterprise is not going to overhaul anything. The enterprise is going to ask for a very small workflow change, if any. They're going to hope that you show up on their existing dashboards, if you show up at all. And they want a small inflection where they can measure an ROI in two months or something. I'm not saying that's good. I'm not saying it's bad. I'm saying that's the reality. And so we started with companies that thought they were going to change everything. And both they didn't know enough, and enterprises weren't willing to wiggle enough, and everything was very narrow. We're now shifting to a place where people have learned some of the hard lessons. People also have realized if you POC forever, you're probably just building technical debt in a thousand dark corners. It's not really transformation. And we're at this juncture where and I think this is actually very critical. I don't see enough people talking about this. We're at a juncture where this buzzword has taken over the world, right? Gen AI has taken over the world. ChatGPT obviously booted all that off. And the venture backed ecosystem, number one, they're less sure of when they get their next round than they were 18 months ago. So before ChatGPT, they were more flush with cash. Now they're less. So less aggressive marketing, right? We see that because that's what they pay us to do, right? We go to market. You want to sell to Goldman Sachs, you got to be able to know who buys AI there and whatever. So that crowd is not going to market as hard, the venture backed crowd. They still are. It's not as hard, right? A lot of three month old CMOs in the venture world. Let's just say that, right? A lot of turnover there. So the buzzword popped off the enterprise wow, this could be amazing. The enterprise employees are, it's accessible to them. They can use ChatGPT. They can use these tools. How are you going to use the anti money laundering, fancy dancy, identification verification thing? You couldn't as an individual person, so they can get experience. And who's already got the relationships and ready to catch that baseball. It's those existing partners. And so those existing partners are already focused more for their own interest and their client's interest in broader land and expand in capability building, not point solution. If I'm a vendor, I'd love to solve one little dark corner problem, do my white glove work once, and just have you pay me every year. If you're a consulting company, that's not how you make your money. You get your money by taking the jackhammer out and starting to redo data infra where it needs to get redone to get to our three year, five year goals. The services ecosystem is here to catch this baseball right when the venture ecosystem, A, not mature enough on the product side, B, don't have enough cash to go to market right now. Not the same way they did 18 months ago. Is that going to lead to more transformation? Yes, because the buyer is more mature, and yes, because the nature of who's catching the damn baseball is already project transformation bent, not point solution band aid bent. So will everybody transform this year? I'm not saying that. Will a lot of money be made in the services world? An unbearable amount will be, and a lot of momentum will start. And who is the next Accenture? Who is the next Cognizant, for example? That's about as good an answer as I've got.

Andreas Welsch:

Thanks for sharing that. I think that's a great perspective of the scene where the opportunities lie within the ecosystem and within the environment at the moment. One of the questions here in the chat that Sachin was asking is, what's the tolerance for failure amongst organizations? Because not all organizations are going to succeed. Bringing their pilots into product. What are you seeing there? Is there a greater level of resilience because over the last couple of years, so many organizations and leaders have made the experience of what is a good use case, what are the prerequisites for scaling this, for doing change management. What's a tolerance for failure?

Daniel Faggella:

There's more tolerance in one level because people now, for the most part, understand that AI is probabilistic, not deterministic. Not everybody understands it, but more people do than did four years ago than did six years ago. They're not expecting turn it on and it should be able to improve X by Y and never make a mistake. Right, that was an expectation for a long time, brother. And I talked to all the vendors that were running into all those problems for all those years. And where there's less of that now. The tolerance, is it monumental? No. The enterprise always leads with defense. The enterprise doesn't want to rip open a new market and launch new products. The enterprise wants to crimp down and shave off efficiencies and, and focus on compliance with government regulations or fraud or something like that. Defense is where the money is spent. And they're still very defensive about experimentation. But I think mindset wise, they're not as ridiculously closed off as they were in part because they understand this stuff a little bit better. But also, yeah, there's still not a lot of wiggle room. There's still not. And, the vendors, especially the service partner folks I've mentioned, they're all beating the very monotonous drum of responsible AI, trustworthy AI, whatever, because what they're trying to signal to the world is, Hey, we'll get you the upside, but we'll make sure that you don't have a PR gaffe where your AI bot is accused of being a an IST a dash I S T of some kind or where you look bad in the press, right? Or you have a project that's a really abject failure. So we'll protect you from hallucinations. We will also deliver those results. And everybody's bringing that message forward because they're appealing to the naturally defensive nature of the enterprise. So is there a lot of it? No, there's not. Has the culture changed enough for there to be slightly more than there were years ago? Abso-lutely.

Andreas Welsch:

In that same vein, if you are then looking to bring AI into the enterprise how, can these leaders really drive them accelerate adoption of AI and Generative AI, knowing that there are certain risks, certain boundary conditions, and that definitely also a varying degree of risk appetite within the organization itself.

Daniel Faggella:

The challenges are what they have been for the last seven years or so since this conversation began to get even minorly mature on some levels. Really some degree of this does have to start at the top. We have to have somebody with enough executive AI fluency to know where this stuff ties to strategy. Because if it doesn't tie to strategy, and you already know this, there is then the only places we would use it would be to solve a bandaid individual problem. We wouldn't use it to invest as a capability, right? If we don't understand it outside of a word. Then I don't want to invest in it. I would only invest in it if it's going to build towards some grander strategic point. What we refer to it as a North star. Where do you want to be in three to five years in terms of how you deliver product in terms of how you serve customers and, deal with customers? So like how you make your stuff, you do your operations and then how you interface with the world, especially the people that pay you cash. What does that transformation vision look like? What's that North star you're moving towards. And then you can think about the strategic levels of leveling up data infrastructure wise, capability wise that are going to lead you there. Some leader has to have some vision of that. It doesn't always start as high up as it needs to. Sometimes it starts with VPs that build enough buzz. And we try projects, we fail projects, we try projects, maybe one or two of them work out, and then eventually the upstairs get smart, but we really need as much executive fluency upstairs as we can get. And then we do need at least some minor degree of consensus around what that end state is. Basically, that transformation vision we're trying to head towards, what is that? And then we can think strategically and say, Great! We don't want to just invest in AI. Oh, this application is hot right now. I don't know if that's the right move for us. This use case demo looks really cool. I don't know if that's the best use case for us. Our competitors did a press release. You and I both know how much that drives a busy activity. Oh yeah, let's go and do that now. There's a better way to make those decisions where we can, yes, pay attention to what's happening in the market, but think about what's an investment? And an investment is building towards a place we want to get to. And so if we think about investments and we can ask, how do we get both near term and long term value along the journey, somebody upstairs has got to build that and then them and the people underneath them have to have some consensus on how to build that upwards. To do this enterprise wide is tough, but that dynamic I just explained is going to happen within departments, within organizations and that roadmap graphic is, like the best way that we've discovered to visualize it. But those fundamentals are exactly the same as they were five years ago.

Andreas Welsch:

So I'm wondering then with that in mind in painting a picture of where do we want to go, where do we want to be in three years and in five years, what do you think? Where's Generative AI going to be in that timeframe? Yeah. And I know you said you don't have a crystal ball.

Daniel Faggella:

I'm willing to throw my hat in the ring and if I'm violently incorrect, I'm willing to have people laugh a little bit at me, but hopefully sympathize a small amount as well. I think that within five years, we would see change. And when I say change, I mean the thing that a big enterprise company doesn't normally do very quickly. They play defense, they don't change quickly. But there's two things that would make that a survival necessity. Again, the mother, of invention is necessity, or whatever the phrase is that's evading me. One is, if new sort of players start eating their lunch or eating up market share in different important ways. That's one. So that's the ground up. The other side is when their peers from the sides start driving astronomically more efficiency or start delivering a product that's better or what have you. That momentum hasn't started with enough of a torrential, obvious impact that anybody feels like we gotta do X. And, on some level, the technology maybe hasn't necessarily been there yet. But I would suspect that there are all kinds of low hanging fruit applications. Let's just say you're talking about five years. Sorry to zoom out. Let me do it. Let's say we pause the technology today. Zero new innovations. No improvements in video models. No improvements in text models. No improvements in hardware. Freeze the tech as it is today. If we go forward five years, I think there's all kinds of LLM able stuff. Where we really do need to rethink how we're handling legal. We really do need to rethink how we're doing this underwriting workflow. We really do need to think and like the FTEs and how many we need and where they get moved to are going to have to be real discussions right now. There's a lot of momentum, keeping those discussions back. Tech's not mature enough for the most part. Also not enough competitors or peers are eating our lunch for us to have to jump. I think there's going to be departments that have to jump. And that's going to mean big shifts in workflows. Actually doing the overhauling that everybody dreamed of eight years ago, seven years ago. And real consequences on employment which I'm not saying are going to be dystopic, but I'm going to say are going to be significant. So if you gave me a five year crystal ball, even with zero tech development, that's what I would say. God knows if we have something as big as LLMs pop off in the next six months. Brother, I'm out of the race at that point. I got no prediction for that one.

Andreas Welsch:

If I'm being asked the question, it's tough to make that prediction when you don't even know how fast is this tech moving in the next three months or six months, let alone in anything longer than a year. So thank you for sharing that. That's awesome. Now we're getting close to the end of the show and I was wondering if you can summarize the three key takeaways for our audience today before we wrap up.

Daniel Faggella:

Yeah, sure. I would say number one, a focus on those critical capabilities I mentioned early on. I really think the receptivity in the boardroom and in the C suite is higher than ever to make that a real discussion. And I think that's going to be absolutely critical. I think that being able to have some kind of shared vision that we crystallize as an executive team around strategically where this could help us position ourselves in the market and what we could become as a company that will permit us to make investments, not just band aids and AI. And again, that, that conversation is rife right now. This is the time to have it. And then in terms of number three, I would definitely encourage leaders to stay abreast of what AI is capable of and where the application space is going. And obviously anybody tuned in right now is doing exactly that. So they're doing a great job by listening to you. So those are my three, my good sir. That's what I got for you.

Andreas Welsch:

Thank you so much. for joining us today, for sharing your expertise with us. It was a pleasure having you on.

Daniel Faggella:

Glad to be here. My pleasure.

People on this episode