← Digital Disruption with Geoff Nielson
Digital Disruption with Geoff Nielson
The 6 Technologies That Will Define the Future with AI Futurist Steve Brown
00:56:46
Loading summary
Transcript107 lines
- [00:03]
Jeff Nielsen
Hey everyone, this is Jeff Nielsen with Digital Disruption. Joining me today is Steve Brown, an AI futurist who's actually worked with Google, DeepMind and Intel in the role of futurist. Steve, thanks for joining me here today.
- [00:15]
Steve Brown
Yeah, pleasure to be here. Nice to meet you.
- [00:17]
Jeff Nielsen
Likewise. So you literally wrote the book on emerging technologies a handful of years ago. The Innovation Ultimatum, it was called, and you covered. Awesome. Yeah, there it is. And you covered six emerging technologies at the time. I think it was 2019, which in hindsight is kind of a fun time to be making prognostications on technological change. And I'm curious, just kind of revisiting that now. Do you have kind of an updated viewpoint on those technologies? Which ones took off, exceeded your expectations, maybe have been slower to adopt know where are they now?
- [00:58]
Steve Brown
Yeah, I wrote the book in 2019. It dropped February 2020, right before. Yeah, it's like really amazing timing for a book. Yeah, it said things that came out of that book. So I mean, the AI was the largest chapter in the book. Yeah, I'm glad about that because it turned out, yeah, that was probably the most important thing to be talking about, Blockchain. You know, I wrote the book with a 10 year time horizon.
- [01:26]
Jeff Nielsen
Right.
- [01:27]
Steve Brown
So it was designed to be, you know, what's going to happen in the 2000s and we're halfway through. So, yeah, I think blockchain is doing what I expected, which is it has some applications in supply chain, but beyond that, I mean, yes, crypto's going crazy, but it's just. Is it creating business value? No, not yet. Which is kind of what I called the things that I thought would go a bit faster than they have autonomous vehicles. Turns out that's a much naughtier problem than any of us expected. And Robo taxis aren't among us quite yet, but it's looking like that will be on track. The thing that I leaned into, that I'm glad I did, and that has turned out to be the most prescient was this notion of process orchestration. Thinking about the looking at the major business processes inside a company, breaking them down into tasks and then figuring out which task is best done by a human, which is best done by an AI, which is best done by a robot, which is best done by an AI and a human working very closely together. I've had some version of a slide that shows that in my presentations for, I don't know, six or seven years now. And now that's the core of what I think every IT person, every line of Business person is having to think about how do you re engineer your business given the capabilities of AI now and where it's going to be in the very near future.
- [02:50]
Jeff Nielsen
So let's talk about that one for a minute, Steve, because you know, I agree with you. I've heard a lot about process orchestration and it completely makes sense to me. I'm curious though, you know, from an adoption standpoint, are you seeing people doing it? Where, where are most organizations on the curve? Are, you know, 80% there or 20% there? And where do we go next?
- [03:11]
Steve Brown
I think it's very early days still. I work with clients in a wide range of industries and they're all panicked because they're behind, they all think they're behind and no one's really that behind. I think people are starting to have to confront some of the fundamental stuff that they didn't quite have right in their organizations in the first place before they can move forward. So where's my data? Is it well organized? Is it clean? For most companies the answer is no, it's not. It's not quite ready yet. So most companies I work with are in that position, but they are starting to map out their business processes and starting to do experiments. So they're doing proof of concept, particularly in, I mean the two main areas are sales and customer support. So partnering a human with an agent that helps them to seem smarter on the phone about your products or services, your trying to sell helps them be better informed about the person they're about to call and what's been happening in their company in the last six months and guides them on maybe what are the right products and services to sell and just automating away the stuff that salespeople hate like filling in CRM forms and having to do those follow up emails. Most of that is now automated by some of these platforms. So those are the areas that I think people are working on first because they're easiest to do. Right? You just build a rag database and pull in your ticket information or you pull in your product information and off you go. Beyond that it gets a little bit harder.
- [04:42]
Jeff Nielsen
And sales can be a particularly sexy place to start because there's this belief that, well, if we start with sales like Cha Ching, the money's going to follow here, it's going to drive revenue. But you talked more about efficiency when we talk about sales. Are organizations, are you finding out actually able to use this technology to improve sales, increase revenues, or is it more of an internal efficiency play? At this point.
- [05:13]
Steve Brown
I haven't seen Any numbers that show me that you're actually increasing revenue per person, but I suspect that that is what is happening because you're just increasing the efficiency of that person. They probably can make more calls per day if the AI is helping them. So a lot of these AIs just listening to a call and then they surface information on the screen for the salesperson, say, hey, you know, there's an objection here. Why don't you pivot your strategy and try selling these? I have to imagine that that is going to give some sort of uplift on the sales process. But I haven't seen numbers yet.
- [05:48]
Jeff Nielsen
Right, right. So, yeah, I guess if you, if you kind of untangle that, it can juice the individual, you know, pieces of the sales equation and you would hope that that would lead to better sales on the other side.
- [06:01]
Steve Brown
Yeah, if you can make more calls in a day. Just kick my light there. If you make more calls in a day and you're able to close those calls more effectively, you should be closing more business.
- [06:13]
Jeff Nielsen
Yeah, yeah. You mentioned blockchain as an area that, you know, it sounds like it's just kind of quietly doing its thing in the background now. It's got its applications, but it's something that maybe doesn't make the same news cycles as it used to. Exempting all the conversation happening now about the price of crypto. Have we reached peak blockchain? Is it no longer an emerging technology or do you see more kind of adoption and innovation around the corner there?
- [06:46]
Steve Brown
There's still potential in the supply chain, a lot of potential. One of the biggest, biggest things that people, whether you're a brand manager, a regulator, a consumer, you want to know where things came from and how they were made. That's. I'm a co founder of a startup, so small plug for the Providence chain network, trying to figure out how to move from a system of claims where I'm selling you something and I claim that it has these features and this provenance. You know, it was a T shirt that's not sewn by tiny fingers in the Vietnamese factory. It is palm oil that didn't come from a dodgy plate. It is leather from hides that were not from cattle where land was cleared in the rainforest. You want to be able to move from a system of claims to a system of evidence. And I think blockchain is really good at helping to do that by locking in this evidence so you can be more confident when you are buying products where they came from. I think that's the biggest and most powerful application the other one is distributed systems being able to create value in a distributed way. And, you know, you're seeing that start to happen with AI, with people being able to run distributed models, basically renting someone else's computer to run models and do training runs. That's an interesting collision between AI and blockchain I'm watching.
- [08:08]
Jeff Nielsen
Yeah, yeah. It's really interesting to me that the supply chain piece, Steve, and the claims piece, and it's so. I don't know, for me, it's just intellectually interesting that, like, literally any object or consumable in this modern world, you know, I can point at this microphone and it's like, where did any individual piece of this come from? Like, I have no idea. Right. Like, I don't know what it's made from. I don't know how it was sourced or what that supply chain looks like. And, you know, the idea that you can have this, you know, this record with evidence of where it came from is, you know, kind of transformative. But I don't know, is this really on top of minds for a lot of organizations or a lot of consumers? How do you cross that hurdle? And what does adoption look like to start putting something like this in practice?
- [08:54]
Steve Brown
It depends on the organization. If you are the US Space Force and you're putting a $100 million spy satellite into space, you need to know that not a single component came from China and you have a room full of people running that information down. Use a product like the Provenance Chain Network, and suddenly you can do that automatically. So if you're a brand manager and you've had maybe a few incidents that have not reflected well in your brand, and they're in the press, you're going to be more interested, or you're just a good brand manager who worries that you're going to be the next person in the press. So those are the sorts of companies that are leaning into this stuff earlier. I think consumers are a little bit further behind. There are plenty of conscious consumers, but there are not enough of them yet. And they're not demanding this stuff enough yet that it's happening. It's being deployed more broadly. I think what may happen is you'll see the next generation of food safety requirements, and they may cause at least the food industry to move in that direction. So they know when there's some contamination in lettuce, as the seems to be on a regular basis. They don't have to take every romaine lettuce off the shelf everywhere in the country. They can trace exactly where it came from and not have to throw all that lettuce away.
- [10:15]
Jeff Nielsen
Right. So, I mean, whether it's lettuce or spy satellites, I don't think I've ever put those two in the same sentence before. It sounds like regulatory bodies may be kind of a key ingredient in this. Do you buy that or is this still going to, you know, kind of proliferate without it being a compliance requirement?
- [10:33]
Steve Brown
That compliance is part of the thing. But I think it's. If you care about your brand, you're going to want to have eyes on your supply chain. I mean, a lot of brands have supply chains that are quite deep. They can go back 3, 4, 5, 6 steps and you can tell your immediate suppliers to behave. Yeah, but you know, economic pressures mean that people are always looking to save a buck here and there. And so two or three steps down the supply chain, you need some incentive mechanism built into your supply chain so that you can incentivize or penalize people if they don't comply with what you're requiring.
- [11:06]
Jeff Nielsen
Right, right. So, I mean, it sounds like overall you're still pretty bullish on this whole use case for blockchain.
- [11:13]
Steve Brown
Very much so, yeah.
- [11:15]
Jeff Nielsen
Oh, it's interesting. I want to, you know, I was thinking about just your list of technologies and one of the ones we haven't really talked about yet is, is VR and AR and just kind of this notion more broadly about spatial computing. And as recently as like a year ago, it seemed like this was kind of on the tip of everybody's tongues. And I don't know if it was Mark Zuckerberg not having anything new to show or Apple having their rollout, but it feels like it's fallen off a little bit. Is this just that point in the cycle? Is it going to come roaring back? What's next in this space?
- [11:53]
Steve Brown
I think with VR particularly, the use cases are relatively limited because when you put on a VR helmet, you're cutting yourself off from the physical world. So in an enterprise situation, you need your workers to go to see what they're doing. So that's why augmented reality mixed reality was always going to be more interesting than virtual reality. Virtual reality is fine for simulations and training for games. It has those sorts of applications. But it's not until you have high quality, low cost, lightweight, long battery life, mixed reality, particularly mixed reality that is paired with an AI assistant that just spatially aware, then it gets to be the kind of use cases that everybody will want, but the technology is just taking a little longer to develop.
- [12:39]
Jeff Nielsen
So you still see it as if we talk about the 2000 and 20s, the next five years, you're still predicting pretty interesting advances in that space once AI can kind of unlock that. Is that fair?
- [12:52]
Steve Brown
Yeah. I mean, you look at what Microsoft is doing with the Dynamic 365 platform and comparing that with HoloLens 2, putting on a pair of magic glasses that then shows you how to fix something or how to build something in the moment just cuts down training times, makes it easier for somebody to get up and do something. And ultimately I'd love something like that to help me build Ikea furniture or fix my washing machine. YouTube videos only go so far to be able to just put on a pair of glasses and have an AI converse with me and talk me through what I need to do is a very powerful use case.
- [13:30]
Jeff Nielsen
Yeah. And there's so many just in this constellation of kind of technologies and, and I use the word constellation deliberately because they all, as you said, they interact with each other. Right. This is maybe an unfair question in its specificity, but obviously a couple of years ago we had that watershed moment with ChatGPT. Do you have any predictions, Steve, for where are we going to see the next one? And is there another emerging technology watershed moment coming within the next two or three years?
- [14:00]
Steve Brown
Yeah, don't worry. I get asked that question all the time. What's the next ChatGPT moment? I think there's a few that we can imagine. I mean, the first time there's an assistant that you can give quite a complex task to and it goes away and does it for you.
- [14:16]
Jeff Nielsen
Right.
- [14:16]
Steve Brown
Which I think is coming next year. And that, that's a pretty big one. The first time you see a robot walking down the street, I think it's going to be a big moment and then, you know, within a year, like, ah, whatever, you know, it's just getting groceries for Lynn down the street, you know. So I think it's going to be moments like that which surprise us all and then they're just going to become normal.
- [14:41]
Jeff Nielsen
Yeah. So robotics, I mean the robots walking down the street and I think I saw like actually in LA or something, it seems like that, like that is starting already. And I'm curious because robotics is not, you know, in innovation Ultimatum Robotics was not one of the big technologies. And so I'm curious, you know, halfway through the 2000 and twenties or the ten year time horizon, are there any technologies that you're starting to look at now that didn't make the cut five years ago?
- [15:13]
Steve Brown
Yeah, robotics was in there as autonomous machines. But yeah, I didn't see the speed of development of the humanoid robots. I've been very impressed with companies like Figure Sanctuary up in Vancouver, bc, Agility Robotics down in Oregon, doing good work and Boston Dynamics are pretty good. So I think we're going to see big leaps forward there. And it's the combination of these robots and advances in motors and batteries, but combining that with developments in large language models and vision systems that's really provided these breakthroughs. What do I see coming next? I deliberately didn't write about quantum as one of the six technologies because I was writing about broad application in the 2000s. I think Quantum is still a 2000s technology, but that's the one I'm starting to watch closely. Is there an intersection between Quantum and the future of AI? We cannot scale up AI in the way that we're scaling it now indefinitely. You're going to hit a wall. We're starting to get there in terms of the amount of power it consumes, the water these things take, you know, it's not environmentally sustainable. So we're going to need something different. And you know, there could be some analog AI stuff. Quantum AI, those are the ones I'm watching closely.
- [16:42]
Jeff Nielsen
So with that, with that application of quantum, you know, it's, you know, I talk to people about quantum, I hear about it, but most of the conversations are still fairly academic in that space, it seems. Around Quantum, are you starting to see any more kind of applied use cases of quantum even, you know, kind of conceptually that are emerging?
- [17:02]
Steve Brown
I mean, again, is why I didn't write about it, because it's out there still. Right though it's interesting in sort of molecular chemistry research looking at the interactions of complex organic molecules, making, doing bleeding edge scientific research. It's a great tool in some specific areas. But are you going to have quantum in your next smartphone? No. So I think there's some interesting areas where you might be able to do AI in a different way. The brute force, matrix multiplication, transformer model, transformer based model approach that we're using for large language models today uses enormous amounts of compute. Are there other ways to do it? And there are some interesting companies out there who are looking at different approaches and they're using quantum techniques at their foundation. Are they using quantum computers? No, but they're using quantum properties.
- [18:08]
Jeff Nielsen
So you mentioned compute. And certainly I don't think it's controversial to say there's a bit of an AI kind of arms race going on right now, especially with kind of the big tech organizations using as much compute as possible to get this next big thing. And we were talking earlier, Steve, about the fact that everybody's behind in adoption right now. And so I'm curious, in your mind, how big is the next big thing? Or do we still have our work cut out for us with the existing technology?
- [18:41]
Steve Brown
That's a very good question. I mean, there's so much opportunity with adopting AI across the enterprise, but doing it right and doing it at the right speed and figuring out the right use cases to go after is the challenge. Right. And what's the foundation you have to put in place? Yes, there are existing technologies that we'll still be able to get some things out of. You know, there's still value in whatever the next rev of the SaaS software is you use. You know, there's even new features. But I think the transformation stuff, transforming the workforce, elevating the performance of your workers and notice I say performance and not productivity. Productivity is a big one. But elevating the creativity of your workforce, the intuition, your workforce, the decision making abilities of your workforce, figuring out how to do that with AI, that is where the real opportunity lies. So maybe completely differentiate versus your competitors. So I think that's where at least the enterprises I'm working with, that's where they're focusing most of their energy. There's some sustaining work. And yes, we need to keep doing the stuff we used to do, but figuring out where to lean in and where to focus, where am I going to run my experiments? What are my pilots going to be? Where are the areas of most opportunity? And that's something I help a lot of clients with.
- [20:10]
Jeff Nielsen
Yeah. And you know, on that, who do you see and what do you see as being the biggest indicators, I guess, of, you know, an organization that's going to be a disruptor versus one that's going to be disrupted.
- [20:26]
Steve Brown
I mean, there's three main areas that people are using AI in their businesses. First one is customer experience. Second is employee experience. Third is operational excellence. Right. How do you elevate your offering, whether it's a product or service, to the user interface, to an existing product? How do you elevate your offering to improve customer satisfaction, improve utility and differentiate in the marketplace? That's a fairly obvious path. The one that I think people are not spending enough time focusing on is how do I elevate my workforce? So that's the process, orchestration stuff we talked about. And you're trying to boost performance, but there's another reason to do that too. We are moving into a period of time when we're going to have not enough people in the labor force. And that is a demographic certainty. The only way you fill that gap is with automation or immigration. And immigration is probably not on the table for at least the next four years or not. Not in the ways it was perhaps in the past. So automation is going to be a significant part of making sure that we can fill the jobs that people have. So when you are competing for that scarce labor, think about, you know, when you're thinking about what's your next job going to be? Do you want to work for a company that has invested in AI tools, services, assistants, so that when you go into work every day, the things that you don't enjoy doing are all offloaded to an AI and the things that you do enjoy doing that you get satisfaction from, you have support to elevate your performance. So you're going to go work for that company? Are you going to go work for the company that's not invested in creating that amazing AI supported employee experience? So just for access to the best talent, I think companies need to think about that very carefully.
- [22:20]
Jeff Nielsen
Right. And just reflecting on that, there's what's in the hands of the organization versus what's in the hands of the employee as well, who's trying to use AI. So for you, Steve, is like a total lockdown of AI in your mind, off the table for organizations. And in some cases, I guess, where there's regulatory reasons to do that, is that creating just an environment where these organizations are making themselves kind of prey to be leapfrogged by someone who's going to take that chance?
- [22:55]
Steve Brown
Absolutely. There are good reasons why people don't want an employee just randomly using ChatGPT, because when you do that, everything you type in is going to the training data set for the next round of models. And it's not good if you cut and paste confidential information and send it to the AI to train on it. So yeah, there are good reasons to do that lockdown. And that was a fair initial response. But that was a fair response two years ago. It is not a fair response now. There are plenty of options to be able to give to expose employees safely to these types of tools without compromising confidentiality. There are ways to interact with Google, with OpenAI, even on an enterprise plan so that you can keep your data safe and to use rag and other techniques so that you can. The average knowledge worker spends 32 days of the year just looking for the information they need to do their Jobs. So taking all of the corporate information and safely surfacing it, making it available through a vector database to your employees is going to save a lot of time and help them be more effective. That's an easy thing to do and companies should all be looking at that.
- [24:13]
Jeff Nielsen
Yeah, and speaking of easy things to do, I mean there's clearly a spectrum here from the low hanging fruit and some of the like, oh, this is easy to adopt right now to, you know, I have to imagine that this, this conversation around like, you know, process orchestration, I mean this is transformative. But I don't think we'd use the word easy for that one if I had to guess. Right. Like that's, that, that's like fundamentally re architecting, you know, the organization. And so, you know, I'm curious, I know you've been doing some AI workshops with various enterprises. What does that look like? How do you recommend they get started and what kind of journey do you take them on?
- [24:53]
Steve Brown
Yeah, so I get hired by lots of companies to help them figure out what's their AI transformation journey. And it usually starts with me going in and talking to their board or management suite, give them a briefing on AI, help them understand what are the opportunities that lay ahead, what are the challenges they should know, how do they avoid the pitfalls. Then they bring me back usually to talk to the next level down. I give them the same kind of spiel and then they bring me in to do the hard work, which is an AI innovation workshop. I come on site for a couple of days and I use the future casting process to help them ideate future use cases for their organization. So they bring people into a room. We get about 40, 50 people together, work in small teams and ideate potential use cases for the future. And then we map them out in time and figure out which are the ones that are priority and what are the. By looking at all of these potential use cases, what are the things that we need to do to get there? Do you have the talent that you need? Do you have the data that you need? Do you need to acquire that data? Is that through a partnership? Is it through M and A? Is it something that you have but you have to clean up? What's the IP that you need to generate? What are the infrastructure requirements as a make versus buy conversation? So I walk them through all of that so, so that they can start on that journey. And even though you might be building a five year roadmap or a three year roadmap, you know what you're gonna do when you go into the office on Monday. How do you start on that journey? So that's probably the most popular workshop I run right now.
- [26:27]
Jeff Nielsen
Right. And you mentioned there make versus buy. And I have a suspicion I know what you're gonna say, but what percentage of the time are you recommending make versus buy and what are the cases you see if any for make?
- [26:44]
Steve Brown
It depends what kind of problems you're trying to solve. You know, if you are wanting to use a generative AI to accelerate your sales team. Yeah, I'm probably going to recommend that you buy that in because building your own large language model from scratch, not something I recommend for people that costs a lot of money and you're not going to be competitive with what you can rent in. Essentially if you are building a model in an oil company to help you figure out optimal gas uplift, so you're forcing gas into the ground to get oil to come out. You know, that's something you're going to do yourself because it's going to give you competitive advantage and that's a model that you are going to be able to create in house. So it really depends on what is it you're trying to solve, what's the problem you're trying to solve when with AI and what's available out there and what is something gives you a sustained advantage versus it's just something that you need as a base capability for the enterprise.
- [27:44]
Jeff Nielsen
So, you know, on that note, do you have any war stories maybe you can share about, you know, really good, you know, AI use cases or ways. You've seen organizations take advantage of this technology that's, you know, kind of caught your attention is like, wow, I never thought of that.
- [28:02]
Steve Brown
It's amazing where it shows up. You know, one of my clients is a uranium mining company up in Canada. When you're mining uranium, you have a spinning head and the rate of spin and the pressure of the water coming out. You vary it over time to get the uranium ore to come out. I'm not an expert on uranium mining as you can tell, but the speed of spin and the pressure changes. You optimize that process with AI and you increase the yield by get ready 1% and you think, well, is that worth having? Turns out, yeah, because that's worth tens of millions of dollars a year.
- [28:44]
Jeff Nielsen
Huh, interesting.
- [28:46]
Steve Brown
So it's looking for that. What are the areas where you can get biggest return? Even if it's a 1% bump by using AI, it's worth doing in some cases.
- [28:56]
Jeff Nielsen
Yeah, yeah. No, it's funny and it's just the 1%, I guess got my attention because it's just taken in isolation, it's like such a who cares number but. Right. You apply it to any big enough sample and suddenly you start to have a real impact.
- [29:14]
Steve Brown
I guess if you're operating at scale and you can go after 10 or 15 one percents, that's going to be transformative to your business results.
- [29:23]
Jeff Nielsen
Yeah, yeah. Are you working with many organizations, Steve, on that kind of, you know, the process orchestration or do you find that's like fair? That's reserved for the most advanced organizations.
- [29:36]
Steve Brown
Right now I'm not. So my role as a futurist is to go in, help them understand the possibilities.
- [29:42]
Jeff Nielsen
Got it.
- [29:43]
Steve Brown
Build that roadmap and then I step out of the way. So I then hand off to other organizations that are going to help them actually get into execution. But they're partners. I mean companies like Celonis, who's a client of mine actually great to work with by the way. They are a company that helps you do that, mapping that map out your business processes, which is the first step to then figuring out which of these tasks am I actually going to automate. Right.
- [30:08]
Jeff Nielsen
And just thinking about the organizations I assume you're working with are not super different from the organizations I'm working with. Which is to say it's probably not the startups that were founded in 2024 and it's probably not for the most part the Amazons of the world. It's that kind of middle of the economy, bread and butter organizations. And I'm curious, from your perspective, how big is the risk right now for these organizations that they're either going to get displaced by founded in 2024 organizations or that their business model or something that they're doing is going to get absorbed by the Magnificent Seven.
- [30:55]
Steve Brown
That's a great question. Most of my clients are medium sized enterprises or big enterprises. So my clients recently have included bank of America, I've worked with Disney, Nike, JP Morgan, lots of big companies. But you're right, the, the question writ large is what's the risk of disruption from upstarts?
- [31:17]
Jeff Nielsen
Yeah.
- [31:17]
Steve Brown
When you have a powerful technology, if you, if someone else is able to harness it better than you because you have the innovators dilemma or what, you have an organizational torpor, you just want to move fast enough. What's the risk? I think the biggest risk is going to be in education. You know, I talk to people in higher ed and they're either la la la la la, I'm not, you know, not listening. You know, you've got teachers who are more worried about their kids cheating using ChatGPT than they are thinking about, well, how do I prepare my kids to thrive in a world of mass automation? How do I give them a differentiated advantage to survive in that world? And how do I use AI and embrace it to amplify my ability to teach them? That's where people need to think and the risk is to your point, okay, I'm Harvard University, I'm XYZ University. What happens when Microsoft Academy or Google University comes along and offers a really great accreditation? And the turning point is when businesses in corporate America say, you know, what a degree or whatever the rating is from Microsoft Academy is worth more or equal to something from a bricks and mortar university, suddenly then they have a real problem. It's hard to charge 2 or $300,000 for education if I can get the same one from Microsoft for two or three thousand dollars.
- [32:55]
Jeff Nielsen
I love that you went there and you know, I do a lot of work with higher ed and it's, it really feels like they're kind of at that pivotal moment right now. They're facing this enrollment cliff due to demographics, due to, as you said, immigration. The model itself is changing. I don't know, it seems to me like we're in this weird kind of liminal space where everyone agrees that the old model isn't really suited to the present, but we haven't found anything yet that's clearly the answer. So I don't know if you are. And to me, Harvard is probably in a better spot than a lot of the next tier or two rungs down from Harvard.
- [33:39]
Steve Brown
Amazing global brand.
- [33:40]
Jeff Nielsen
Yeah, because they've got the brand. But how do you avoid watching from the sidelines? I mean, do these organizations have an opportunity to disrupt themselves to partner with the Microsoft academies of the world? Where do we go from here to make sure that we're not just kind of in the dustbin of history?
- [34:02]
Steve Brown
Yeah. So I think there is a huge partnership opportunity because the benefit of going to university is not just information transfer from faculty brains to my brain, that's a big part of it. And learning some of the knowledge and skills that I need to thrive in the workplace in the future. But it's also about learning to be a well adjusted human to interact with other people. There's a huge networking component. Right. I still am in touch with people I was at university with and sometimes we help each other out.
- [34:36]
Jeff Nielsen
Yeah, right.
- [34:37]
Steve Brown
Certainly if you went, if you went to Harvard, it's amazing. Network there, that's a big part of why you go to Harvard. It's not just a brand and the cachet. So I think the opportunity is how do you have this blended model? And it comes back again to process, orchestration. What are the things that I can deliver best with an AI? An AI tutor that adapts to my style, that is very good at teaching me anything I want to learn. And pair that with the human connection of a tutor being on campus with people who are like me. And we're learning, we're talking about the ideas that we are being exposed to and building that network for the future. How do you combine those two? So maybe I can do a degree in a year instead of three and get the same high quality education, the same networking and pop out the other side and at a fraction of the cost.
- [35:30]
Jeff Nielsen
Right. And you've got, you've got the gears turning for me right now, Steve, because you know, I see a few things coming together and one of them is, you know, I'm, I've got a business background, but I'm a self proclaimed apologist for like a liberal arts education. And you know, I've always believed that one of the main benefits to schooling is like it teaches you how to think, if that makes sense. Right. Like it's not necessarily about the answers you learn there, it's about asking the right questions. And the gears you got turning in my brain are like, to me, in the age of AI, that's actually more important than ever. Right. Like that the answers, like it happened when we had calculators, like you don't need to remember how to do long division. And something similar may be happening in AI. So I don't know, where does that kind of take us and does that become a skill that's more needed in the workforce? Is that something you're seeing?
- [36:28]
Steve Brown
Yeah, yeah. I mean there was an article, I think it was in the Atlantic about a year, year and a half ago, which the headline was the most Learning to talk to an AI could be the most important skill of the 21st century. They are going to be the tools that we use to help us do all the things that we do in our personal lives, our private life, work lives, learning constantly growing AI is going to be a partner for doing that. And the same way we all had to learn, if you're old enough, WordPerfect and Lotus 1, 2, 3, if you're younger than that, then you know, Word and Excel or the Google Suite, you had to learn how to get the most out of those Tools we have to learn to get the most out of our AIs. And that is going to happen through a process of trial and error and being taught. And yeah, I think we're going to need to teach kids how to get the most out of their AIs. But at the same time, people always ask me the number one question I get, probably number two is the one you asked me about, what's the next chat GPT moment? The number one question is some version of what do I need to do to become robot proof? Or what do my kids need to do? What do my grandchildren need to do? What should they do in school to make sure that they're robot proof in the future, that they're going to be relevant in the workplace? And I tell them the same answer every time. Double down on your humanity, the things that make you uniquely human. Your empathy, your communication skills, your intuition, your creativity. AIs can do some creative things, but humans are better for a while. So figure out what are the things that make you uniquely human. Lean into those, because the future is still human. It's just supported by amazing AI.
- [38:20]
Jeff Nielsen
So thinking about, I love the optimism that comes with that. And I was going to ask you, on balance, what is your excitement versus fear, looking out at the next 10 years and maybe more than anything, what it means for children and grandchildren and the future of somewhere between humans and the human workforce, I guess.
- [38:45]
Steve Brown
Yeah, I mean, we're getting into a huge question space now.
- [38:50]
Jeff Nielsen
Guilty.
- [38:51]
Steve Brown
You know, if you ask the question, you know, should I be excited or afraid of AI? The answer is yes. You know, it's both. We need to be cautious about it. It is a very powerful technology. It is an amplifier, so it amplifies good behavior and bad behavior. So we could live in a world of huge misinformation and disinformation created and pumped out from AI factories at scale. At the same time, there's a good chance that somebody listening to this podcast will have their life saved by an AI. A drug that was developed or co developed by an AI is going to save someone's life, if not you, then your kids. So it's good and bad. My biggest worry, I'm a future boy, right? So when I play things out, I don't know how long it will take, but AI is probably going to get good enough that it can do most jobs. Most economic activity will be done by machines, freeing us up to do something else. And I'm not sure that we all know what to do with ourselves. And let's say we can fix the economic side of things and make sure that people have what they need to live. If we live in a world of abundance fueled by AI because the cost of goods mostly is about labor. So if you crash the cost of labor, then the cost of goods comes down. You live in a world of abundance. Everybody can afford what they need. So you figure out some sort of economic distribution model, then what? What are we going to do with ourselves? And so I think we all have to start asking and answering questions to prepare ourselves for whenever that comes. And maybe it's a generation from now, maybe it's two generations from now, maybe it's never, maybe it's 15 years, I don't know. But when it comes, we have to know who are we? What's important to us and what makes us happy. And it's never too late, never too early to start thinking about the answers to those questions. Now, we all found out during COVID when, you know, the universe sent us back to our bedrooms to think about what we'd done for a few months.
- [41:06]
Jeff Nielsen
Yeah, yeah.
- [41:07]
Steve Brown
People turned to alcohol and drugs and they were miserable. When that happens and you're told you don't need to work for the rest of your life, what are you going to do?
- [41:16]
Jeff Nielsen
Yeah, well, that's my biggest worry, is.
- [41:20]
Steve Brown
We won't make that transition elegantly.
- [41:22]
Jeff Nielsen
Well, and that's. There's something implicit there too, Steve, which is to what degree we answer that question, ourselves versus someone answers that for us. Right. And in some ways, maybe that was the scariest part of the pandemic, is that there was no one there to answer that for us. Right. Because we didn't know. And it feels to me like we're just at this kind of historical inflection point where there's just more unknown unknowns than ever. And the number of people who can tell us what's next seems to be dwindling. You and I try our best, but it's just, I don't know, I feel like we're still in kind of watch and see mode as we kind of grapple with these bigger questions. So I don't know, should we be waiting for somebody to tell us what's next, or should we be empowering people to come up with those answers by themselves?
- [42:14]
Steve Brown
I think we all own this. Right. Our society is constantly changing. We're about to see more change in our society in the next 20 years than perhaps the last hundred, perhaps the last 200. You know, that's a big deal. And we all need to help each Other through it. And yeah, I think businesses will pop up that help people figure out what's your life 2.0. Right. There'll be people to help with that. But I'm encouraged by the fact that, yes, some people turn to alcohol and drugs and didn't do so well and came out of COVID depressed, and I feel very bad for them. But there are also people who figured out a love of making bread and getting out in nature more and discovered arts and crafts that they've carried on to this day. So I think that that little glimpse that two or three months where we were all in lockdown was a test. And some didn't do so well in it, and some did. So we need to help people who didn't do so well in that as this becomes something we face on a grand scale, help them make that transition.
- [43:20]
Jeff Nielsen
Yeah. And with that transition, do you see AI as a risk to people finding meaning? And what I mean by that is there's this kind of talking point going around now that, you know, AI was meant to do all the administrative stuff so that people could do all the creative stuff. And lo and behold, suddenly AI is writing songs and it's creating paintings and, you know, it's people who are left doing the dishes, which AI can't do. Does that have merit? Or do you see that as just kind of, I don't know, an easy talking point that won't bear itself out in the long term?
- [43:57]
Steve Brown
I mean, it's a fair critique of AI today, but at the same time, people who don't have the ability to draw amazing images, but they have an idea in their head that they can't realize and they've been impotent up to this point. Or people who want to write a song and they know what they want, but they can't sing, they don't play an instrument, now they have access to that. AI is going to help them realize their dreams. So I think in some ways it is unleashing people's creativity. They're just using these tools to create the output rather than picking up a guitar. Or some people will argue, well, if you don't have calluses on your fingers, you're not a musician, you're not creating music. Well, there's some truth in that, but not everybody wants to take that path. And I suspect that we will have. People will want artisanal things, things that were made by humans. It doesn't mean there won't be a marketplace for things that were made by machines guided by humans. So I think you'll see both. There's a place for all of it, I think.
- [45:05]
Jeff Nielsen
Yeah, it just becomes kind of a two speed economy or a two tier. There's the. Are you willing to pay the premium for the bread made by human hands or whatever use case you want to use? One of the, one of the other potentials I see here is when we look at what AI is able to unleash for people, there's the creative side that we discussed, but there's also the productivity side and there's also the ability in my mind of if people want to start their own enterprise, their own company, the ability to do that now seems like it's an order of magnitude easier than it was 10 years ago. And it seems like to your point, Steve, it's only getting easier. I mean, is this the direction we're going in? And is that something that enterprises of 50 or 100,000 people should be worried about if suddenly any individual can run their own organization and play in their space?
- [46:13]
Steve Brown
I mean, it's empowering of change. Jensen Huang, I mean, he's running the most valuable enterprise on the planet right now. And he famously was asked a couple months ago, you know, what's your plan for the future of Nvidia? And he said, well, we've got about 32,000 employees now. We're planning to head to maybe 50,000 human employees supported by 100 million agents.
- [46:38]
Jeff Nielsen
Wow.
- [46:38]
Steve Brown
And that tells you what the workforce of the future is going to be. It's going to be humans, digital employees and robots working side by side. You know, and people are thrown around the idea of, you know, when are we going to see the first unicorn, first billion dollar enterprise with one employee? You know, it's possible, I think it's more likely you'll see a group of a few, you know, basically a C suite supported by agents. But yeah, it's, it's going to unleash an enormous amount of opportunity and creativity and that's a good thing. But yeah, if you're a big enterprise that's kind of stuck in the past, the average life of a company is what, 40 years? So maybe it's your time, which means you need to start thinking now about how are you going to scale your operations using digital employees. Employees who work for electrons, not dollars and cents.
- [47:31]
Jeff Nielsen
Right. And just on the point there that you mentioned about, oh, maybe it's a C suite and then down from there it's. AI was going to use the word risk. I don't want to use the word risk. Do you see us Moving toward a world where there's AI C suite, where we get to an AI CEO, or is that too fanciful?
- [47:55]
Steve Brown
I don't think it's too fanciful. It's not going to happen right away. It starts off where as these digital agents start to come on the scene. And OpenAI is rumored to launch their operator agent in January 2025. So that's when we'll start to see this happen. We're going to move, we'll move quickly into a world of 100 billion software agents, which means that every one of us will be a manager. We will all manage really smart interns that we can have. They act like remote employees. But the question is, when do we then start to be managed by an AI? In some circumstances, an AI may be more capable as a manager. Will they move to CEO level? Once we get to AGI or asi, maybe all bets are off. And yes, we will have an AI that is going to do a better job than a human. But will we accept that is the real question.
- [48:58]
Jeff Nielsen
And I'm glad you brought up AGI because that's another question I'm sure you get fairly regularly. Is, is AGI something we're going to see in the next five years and for a lot of the conversation we've had today, Steve, like do we need AGI to get to these places or is it just one, One potential, one potential innovation?
- [49:22]
Steve Brown
Yeah, I don't know if you need AGI to get to these places. What you need is a highly accurate AI. The accuracy of AI is not there with hallucinations. There are some techniques you can do to improve that. The reasoning capabilities you see in models like O1 01 Pro, particularly using rag so that you're operating off a source of ground truth help a lot. But reasoning and RAG stuff only get you so far. It needs to be more accurate than it is today, so it's reliable once you get to that level. I mean, these AIs are already operating at sort of PhD level. You know, a digital PhD employee that I can hire for 50 cents an hour because I'm just paying for electrons. Yeah, that's interesting. I don't think it needs to be AGI to do that. When does AGI happen? Sort of implicit in your question. The smart people I speak to in the industry, the range I get is from next year to never.
- [50:22]
Jeff Nielsen
And it's somewhere in between nice and nice and precise.
- [50:25]
Steve Brown
Yeah, but the smart people, you know, I'm hearing 2028, 2029. Ray Kurzweil bless him, as has been saying 2029 for 40 years. It would be kind of nice if he was right. Yeah.
- [50:39]
Jeff Nielsen
Interesting, interesting. What are you seeing people get wrong about conventional wisdom around AI right now?
- [50:50]
Steve Brown
I think the adage that Bill Gates uses is right, which is people always have. They always overestimate the impact of a technology in the short term and underestimate it in the long term. I think that's the most likely with AI. There's a lot of fraud about AI right now. Everybody's excited about it. And, you know, you had me on this podcast because everybody's excited about AI, and rightly so, but we're probably going to see some sort of a dip coming as people's expectations fall, you know, fall short of what's delivered. But in the long term, I think we're going to see a wholesale transformation of our society because of AI. But that's probably the best answer. Are there specific things that. I think there's just so much hype and excitement over AI, and it's well placed, but it's just going to take a little longer than people think in the short term for it to have true impact in the enterprise. But then once it starts to find its way in, I think you're going to see dramatic change.
- [51:49]
Jeff Nielsen
Right. So, you know, thinking about that dramatic change, Steve, you know, you mentioned before we started recording that you're actually working on a new book right now kind of revisiting, you know, what the future looks like from an emerging tech landscape. Can you maybe give us a sneak peek about what's in there, what you're excited about, and what we should be thinking about over the next five or ten years?
- [52:11]
Steve Brown
Sure. So the first book I wrote was called the Innovation Ultimatum, and the premise of that was to. To arm business leaders with the right questions they needed to ask so that they understood the technology at a level where they could work with vendors and figure out how to use this basket of six technologies to innovate in their organizations and the imperative to do that, hence the Ultimatum. The second book is going to be creatively titled the AI Ultimatum, and it is really leaning in on that one technology and trying to do roughly the same thing. So help business leaders, equip them with the right questions, to ask some frameworks to help them think through how and when might I use AI and what kind of areas might it help me move my business forward and then empower them to participate and build a roadmap for an AI transformation for themselves personally and also for their organization. So that's hopefully coming out next year.
- [53:11]
Jeff Nielsen
Amazing. Looking forward to reading it and diving into that. Steve, what technology or implication of a new technology is most keeping you up at night these days?
- [53:23]
Steve Brown
Ooh. I mean, other than AI potentially challenging our role in the world and giving us a lack of purpose. That's.
- [53:34]
Jeff Nielsen
That one.
- [53:34]
Steve Brown
Yeah, that's. That's a long term worry and I think we will rise to it if we start planning for it now. But that's a long term worry. One of my most keeps me up at night. I just, I worry that we're not ready for the speed of change that's coming. It's going to come at us so quickly and I think we've already seen over the last two years just how quickly the AI roadmap is moving forward in terms of its capabilities. It should prepare us somewhat for the speed that things are going to run. But I just. Once this starts to bite and it starts to be deployed broadly in enterprises and that has implications for the workforce, I'm not sure that as a society we're ready to handle that speed of change. That's my biggest worry. But if we can start talking about it now, then I think we have a good chance.
- [54:31]
Jeff Nielsen
Right. So how we make sure that it doesn't just happen to us, which is the risk right now that we're just.
- [54:39]
Steve Brown
Yeah. I did a ted talk in 2012 or something. I think I still had a bit of hair then and the title of my talk was why Machines Must Make Us Better Humans. I think we need to have that mindset of rather than figuring out how to use machines to automate away talent, how do we use it to amplify and elevate human talent? And I think that just, that makes more economic sense in the long term, but it also is going to enable us to be all much happier in what we do. It's not all just about a bottom line and making short term profits. I hope that people will figure that out.
- [55:24]
Jeff Nielsen
Yeah, yeah. It's, it's a lot to unpack and yeah, it's, it's going to be really interesting to see where the next few years go.
- [55:33]
Steve Brown
Yeah. I mean, buckle up. 2025 is going to be exciting year.
- [55:37]
Jeff Nielsen
Yeah.
- [55:37]
Steve Brown
And I think 2026 beyond that.
- [55:40]
Jeff Nielsen
Absolutely. Steve, do you want to do a. Just grab a sound bite of doing a quick plug for your website?
- [55:46]
Steve Brown
Sure. So easiest way to find me is at my website, SteveBrown AI. I do a range of different keynotes for public audiences for C Suite and everything in between. I do AI innovation workshops to help people figure out use cases of the future for AI. And coming soon will be an online training course on all things AI designed for business leaders and anybody who's interested to learn about AI on how you can use it to actually build your AI transformation journey. So stay tuned. Coming soon.
- [56:15]
Jeff Nielsen
That's great. Yeah. Anyone we can get ahead of the curve and not just slapped in the face by this thing is is a win in my books.
- [56:23]
Steve Brown
Yeah. Make the future happen because of you, not have the future happen to you.
- [56:28]
Jeff Nielsen
There you go. Steve, thanks so much for joining today. I really appreciate it and really, really insightful conversation.
- [56:34]
Steve Brown
Yeah, we enjoyed talking with you. Thanks for the great questions.
- [56:37]
Jeff Nielsen
Amazing. Thank you.