The AI Daily Brief: Artificial Intelligence News and Analysis - What Most Companies Will Get Wrong About AI At First

Episode Date: October 10, 2024

NLW joined Nationwide at their TechX event in Columbus, Ohio yesterday. On this episode, he gives podcast versions of five discussion points from the keynote. Concerned about being spied on? Tired of censored responses? AI Daily Brief listeners receive a 20% discount on Venice Pro. Visit ⁠⁠⁠https://venice.ai/nlw⁠⁠⁠ and enter the discount code NLWDAILYBRIEF. The AI Daily Brief helps you understand the most important news and discussions in AI. Subscribe to the podcast version of The AI Daily Brief wherever you listen: https://pod.link/1680633614 Subscribe to the newsletter: https://aidailybrief.beehiiv.com/ Join our Discord: https://bit.ly/aibreakdown

Today on the AI Daily Brief, five questions on the future of AI in the enterprise. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI. To join the conversation, follow the Discord link in our show notes. Hello, friends. As I mentioned yesterday, I am currently on the road. And today, I am actually in Ohio for Nationwide's TechX event. This is an annual conference that's for the nationwide team where they check in on new technologies, discuss how people across the company are thinking about new categories of technology, implementing those technologies to get value, what challenges they're facing,

all of that sort of shared exchange with the goal of team members leaving both more inspired and more empowered to put new technologies into practice to improve the work they do, both individually and as a collective. As you might imagine, given that I am here this year, the theme is generative AI. And it comes at an interesting moment. It is absolutely undeniable that workers, particularly knowledge workers, are rapidly adopting generative AI. It is also undeniable.

That said adoption and just in general the AI transformation is a huge priority for basically every boardroom in the world. And yet at the same time, a lot of companies are now through their first phase of experimentation and running up against a similar set of challenges. Some of those challenges are structural, some of those challenges are technological, some of them are organizational.

And so it's a really great time to be asking publicly, I think, not just within the context of one's company, but more broadly, how different people are navigating, those challenges, getting the most value out of AI, and moving from what can sometimes feel like pilot or proof of concept hell into actually scaling AI systems and solutions. Tomorrow, we will be back with a normal episode. However, for today, I'm going to take five of the questions from this keynote fireside and basically give you a version of my answers here. I'll be interviewed on stage at the event, and so we've prepared a bit about where we want the

conversation to go, and that's what I'm going to be extracting from. Particularly for those of you working inside big companies, this might be really interesting. All right, so first question is around how big a deal AI actually is. I'll have this AI voice pose the question, and then I'll answer it. There is a lot of conversation on the impact of generative AI, ranging from miraculous to treating it as a fad. As someone who spends time every day studying the changes of generative AI and its impacts, how do you think about its potential? Do you see it transforming industries in everyday life?

My answer here in short is that pretty much no matter how bullish one is, how much they think generative AI is going to change the shape of our daily lives, I think that almost all of us are still underestimating it. The impacts are going to be, of course, both in our personal lives and our professional lives. In our personal lives, these technologies are going to transform how we get information. Think about a world pre-Google. Whenever you run into movies from 20, 25 years ago, it's always wild to see entire plots that are predicated upon problems that technology is just completely solved, like not knowing the answer to a particular question, which of course a Google search can find now, or not having directions to a place. And I think it's going to feel similarly

a decade out from here when we think about a time before we had generative AI. There are going to be these fundamental interactions with the world that seem impossible now that are completely commonplace in the future. How we interact vis-a-vis language, I think, is a great example of one. I think travel is going to be a totally different experience, given that translation capacity. And then there are a million other ways that in our personal lives, these things are going to be hugely impactful as well. The transformation and how we learn is going to be the subject not just of entire podcast, but of entire industries. And broadly speaking, there's just going to be a mass expansion in what we can create and how we interact with the world. Now, of course, in our

professional lives, we're already seeing the impact of generative AI right now. It's proceeding through making tasks more efficient, to changing how we work as teams, to changing in a more fundamental way what we think we can do. And this, of course, is where a lot of companies are and where they're getting stuck. It's much easier to start to see from the bottoms up how individual workers or employees can get value and achieve productivity gains. It's much more complicated to figure out how to move that value accrual up the chain from individuals to teams and from teams to departments and from departments to the organization as a whole. And yet that process is inexorably happening. Now, to the extent that those of us who are bullish about generative AI are

underestimating anything, I think it's potentially how complicated this transformation process will be. For almost everyone who interacts with them, these technologies are self-evidently powerful. Part of the reason that there's such a broader set of early adopters than there have been in the past is that it doesn't take much to convince people that these are tools they want to play around with. That said, the applications of these technologies, how to integrate those applications into existing workflows, are much less self-evident. There's a lot more complication. There's organizational inertia. There's human process inertia. So there's going to to be a lot of work there.

This is an area that moves incredibly fast. Things we thought would be years away have become possible in months. Is there anything recent or upcoming in the field that you found groundbreaking? There's that famous Arthur C. Clark quote that sufficiently advanced technology is indistinguishable from magic. For many people when they first use chat GPT or mid-jury, that was the feeling they got. And recently we once again had an example of that where I've seen more commentary to the effect of this is the most magical I felt technology be since ChatGBTGPT, than I've seen in the couple years that I've been covering generative AI.

And that is around Google's Notebook LM, specifically the podcast creation feature of Notebook LM. Now, Notebook LM has been available in some form, at least in beta, for about a year now. However, recently, they introduced a feature where they can turn any document or set of documents into a conversation between two AI-generated podcast hosts. The quality is incredibly high from the standpoint of

the naturalness of the conversation, to the way in which it turns a summary of key points into an interesting and engaging conversation. And I think it's one of those lightning bolt moments that has people thinking about how totally different the future is going to look than the present that they currently inhabit. Now, one interesting concept that I want to introduce here is Jevon's Paradox. Box CEO Aaron Levy tweeted about this last night, saying the least understood yet most important concept in the world is Jevon's paradox. When we make a technology more efficient, demand goes well beyond the original level. AI is the perfect example of this. Almost anything that AI is applied to will see more demand, not less. So basically the idea of this paradox is that when you increase

the efficiency with which a resource can be acquired or taken advantage of, right? So you make roads better or you increase the efficiency of batteries. The paradox is that that increase in efficiency actually increases demand as well. So you make roads better and more people use them. I think notebook LM is a great example of how this is going to play out in AI. A natural question when people see something that looks like a proxy for something they have now is whether the AI thing is going to cannibalize the existing thing. So in the case of Notebook L.M., is it going to kill traditional podcasts? My take on this is that it's going to be one of those Javon's paradox situations, where not only does it not cannibalize existing podcasts, but instead it wildly increases the

amount of audio consumption that most people experience. I think all of a sudden what a podcast means is going to change. People are going to make themselves podcasts for every single thing they have to study in school. It'll be the starting point for almost anything they want to learn, whether in school or out of school. Creators who never would have started their own podcast now will curate voices and conversations and topics and create the podcast that they wish someone else had created already. And as people get more used to listening and consuming audio in that way, they'll naturally expand their horizons and look for other types of similar content. I think that this is going to play out across a million different domains inside the context of AI.

And it's one of the biggest reasons that I'm not bullish going into the future. Today's episode is brought to you by Venice. Venice is a private, uncensored generative AI app. It accesses open source models to enable text, image, and code generation without the fear of being spied on or having your data exploited. Discuss anything with Venice without concern about it being monitored, sold, or given to advertisers and governments. Venice is different because your conversations and creations are kept securely within the browser, never stored or accessible by Venice. Unlike other AI apps, Venice won't tell you what's okay to say or not. Venice won't patronize you. It simply provides direct access to machine intelligence, no topics are off limits, no ideas,

or taboo. With Venice, you're in control of the AI as you should be. Pro subscriptions are available for $49 a year or $8 per month. AI Daily Brief listeners receive a 20% discount on Venice Pro. Visit venice.a.i slash NLW and enter the discount code NLW Daily Brief. That's NLW Daily Brief, all one word. Today's episode is brought to you by Super Intelligent. Every single business workflow and function is being remade and reimagined with artificial intelligence.

There is a huge challenge, however, of going from the potential of AI to actually capturing that value. And that gap is what Super Intelligence is dedicated to filling. Superintelligent accelerates AI adoption and engagement to help teams actually use AI to increase productivity and drive business value. An interactive AI use case registry gives your company full visibility into how people are using artificial intelligence right now. pair that with capabilities building content in the form of tutorials, learning paths, and a use case library, and Super Intelligent helps people inside your company show how they're getting value out of AI, while providing resources for people to put that inspiration into action. The next three teams that sign up with 100 or more seats are going to get free embedded consulting.

That's a process by which our Super Intelligent team sits with your organization, figures out the specific use cases that matter most to you, and helps actually ensure support for adoption of those use cases to drive real value. Go to B-Super.a.I to learn more about this AI enablement network. And now back to the show. In our view, 2023 was about experimentation and understanding this technology. 2024 was focused on the value it can provide. What do you think the theme of 2025 might be for generative AI and how should we prepare for it? Okay, so the question is, if 2023 was about,

experimentation and 2024 was about understanding value and return on investment. What's 2025 going to be about? A couple follow-up thoughts. First of all, I think this is both linear and non-linear. There is a linear answer here, which is basically that if you've gone through a period of experimentation and a period of trying to understand value, the next thing is scaling what works. Like we were talking about before, trying to move value accrual up the value chain from individual employees, seeing productivity gains, two teams being able to design their work more efficiently to the company as a whole. The reality, though, is that rather than viewing these simply as a waypoint in time that gets completed and moved on, like grades in school, the analogy isn't, this is freshman year and

sophomore year and junior year and senior year. Instead, this is a category of processes that have a relationship with one another that are going to get repeated over and over. What I mean is that we will perpetually be cycling through an experimentation phase, an understanding value phase, and a scaling what works phase, just with new waves of AI technology that come online. The obvious next thing that will require some version of this same process is agents. And indeed, this is part of what makes it so hard to be an enterprise figuring out AI adoption right now, is that just as people are starting to feel like maybe they've wrapped their hands around the assistant paradigm, or at least are comfortable playing around in that paradigm,

this totally different thing is coming down the pipeline, which represents a whole new type of interaction with big implications. Agents will once again require, an experimentation phase and one that probably goes on for quite some time. That experimentation phase will eventually give way to an understanding value phase, which will eventually give way to a scaling what works phase, but by then there will be some totally new category of agent, something that's even farther on open AI scale towards AGI that will require this process once again. And so while yes, it is good to think about these things as a sequence, it's also important to understand that that sequence is going to be repeated over and over and over again. Building on the rapid advancements in generative

AI, what impact do you foresee it having on enterprises over the next five years? What are some of the significant changes we should be preparing for? I think broadly, there is going to be a mistake that many, many companies make when it comes to generative AI. And that is seeing AI as a technology that is all about doing the same with less. In other words, seeing generative AI as an efficiency technology that is about cost cutting. There will be a temptation and a pressure, frankly, from public markets. And maybe even a short-term reward for cutting costs. Certainly, you see this in the PR value of companies like Klarna that are talking about how they're able to greatly reduce staff and they're rebuilding big expensive SaaS platforms. If nothing else, they're figuring out how to generate headlines with this.

And like I said, markets are very short-term. They like when companies cut costs, but are still able to deliver the same value. And so a lot of CEOs will have really good years looking really smart, just treating AI as a way to do the same with less. However, the entire history, of the human experience suggests that the real winners in this transformation will not be the people who view AI as a same with less technology, but as a more with the same technology, or even a much, much more with just a little bit more technology. And so what I mean by that is that rather than thinking about AI as a way to cut costs and just do the same amount of stuff more efficiently, I think the smart companies are going to totally reimagine how much value they can deliver

with either A, the same amount of input resources that they are currently deploying right now, or B, by smartly increasing incrementally the resources that they're deploying against whatever it is that they're trying to do, they can unlock totally new types of value creation that simply weren't possible before. The winners in AI, in other words, in the medium and long term, will be the companies who view it as a chance to offer something better, bigger, different, than anyone else, potentially something that simply wasn't possible before. It'll be the companies that turn all of their employees into supercharged versions of themselves, that figure out how to get those employees deploying AI in teams

so that their whole is greater than the sum of their parts, and generally who have a bias towards doing more, better things, rather than being content with the same. I think that most companies will go through a process of first trying to do a same-with-less stage before they figure this out and move on, but there's an opportunity right now for companies to skip the same with less stage and go right to them more with the same stage.

And I think that ultimately those companies are going to be the ones who really shape the future. In your experience, how are different companies approaching the integration of generative AI? Are there distinct strategies or mindsets you've observed and what impact do these have on their success and innovation? I think ultimately there are three pieces of this. And of course, this is greatly reductive,

and it's much more complex in practice. But first of all, one part of integration is around intentionality of a culture shift, getting people to be bought in around these technologies and actively wanting to integrate them. This is an employee level thing where employees are encouraged to experiment and share what's working, but it's also an executive level question where employees need to see that executives are actively thinking about how to use AI to change what they do as well. There has to be an intentionality of culture shift at the base of this. Otherwise, there's just going to be tons of challenges throughout the whole process. Next, I think this interesting,

requires new systems. And by systems, I mean new organizational processes. It's one thing to just encourage people to try AI and share what they're doing. It's another to actually create processes by which they know what things they can try, how they're supposed to share those results back, how promising use cases get scaled up. Basically, orgs need to figure out the processes by which AI can intersect with the actual day-to-day of how they run their company currently. And lastly, on top of those new systems, there's a role for new structures. By new structures, I mean things like actual software that helps aid these processes, which is, of course, what super intelligent is playing around in. Our AI enablement software is all about helping companies better track, share, and amplify the use cases that their employees are engaged in.

We're trying to take something that is currently invisible, which is all of this employee experimentation with AI, and make it visible, so that employees can adopt AI more effectively by copying what's working from others, rather than every single person having to be a little incubator lab. experimenting all on their own. We also think that visibility into what people are doing is going to provide a critical data set for decision makers to figure out how to properly think about strategy around AI, be that AI strategy itself like procurement of AI tools or broader business strategy where AI has a role to play. And of course, super intelligent is going to be just one of a number of different structures that come online that support this broader AI integration. But again, thinking about this in terms of the intentionality of culture shifts, new systems and organizational processes, and new structures to enable those processes,

I think is where successful companies are going to find themselves. So, friends, that is what we are talking about today with Nationwide, like I said, at their TechX event. Let me know how this relates or doesn't to your company, your lived experience. Use the comments on Spotify or YouTube. And like I said, tomorrow we will be back with a normal episode catching up on the AI news. Until then, peace.