Skip to content
S4E1 Organizational AGI episode artwork

Jul 25, 2024 1 hr

S4E1 Organizational AGI

from Invisible Machines podcast by UX Magazine Hosted by Invisible Machines

How do AI agents fit into a strategy for increasing a business' longevity? Imagine having a conversational operating system for your organization that was always learning. The advanced orchestration of AI agents can create a broader technology ecosystem that sets a trajectory for organizational AGI. Whereas artificial general intelligence, or AGI, refers to a system of artificial intelligence that meets or surpasses human capabilities across a whole range of cognitive tasks. organizational AGI seeks to create the general intelligence it takes to run vast portions of a business. Robb Wilson, CEO and co-founder of OneReach.ai, coined the term OAGI and is joined by Josh Tyson, andExec Producer Elias Parker for a forward-looking exploration of organizational AGI. Create and orchestrate conversational AI agents: https://onereach.ai/ai-agents/?utm_source=soundcloud&utm_medium=social&utm_campaign=tanmay_bakshi_episode&utm_content=1 #AI #AIagents #OAGI #AGI #generativeAI #LLMs #UX #AIproducts #conversationalAI #AIforgood #Skynetnotyet

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

S4E1 Organizational AGI

0:00 1:00:37
of matches

Transcript Auto-generated, may contain errors

Welcome to Invisible Machines, a podcast about the expanding world of AI agents and conversational technology. I'm Josh Dyson, co-author of Age of Invisible Machines, a best-selling book all about the agentic approach to automation and the pursuit of organizational AGI. I'm joined each week by Rob Wilson, CEO and co-founder of OneReach.ai. In broad terms, artificial general intelligence, or AGI, also sometimes referred to as singularity, is talking about a computing system that's able to meet or exceed human capabilities across a wide range of cognitive tasks. So it's general intelligence that allows humans to adapt and to thrive in all sorts of harsh environments here on planet Earth. And I guess it probably goes without saying that replicating that level of intelligence in a computing system is incredibly difficult. However, creating a level of general intelligence required to run a business is far more feasible. And that's what we're here to talk about today on episode one of season four of Invisible Machines.

So we're in an era now when businesses can essentially have operating systems for their entire organizations that are conversational in nature and that are always learning. So employees and customers can ask the OS questions and it can connect AI agents and data to provide useful information or to take action across a wide range of shareable skills. Because this OS can communicate across departments and channels, contextualizing unstructured data and learning from its interactions, it can actually become a new type of OS entirely, organizational singularity. This is meaningful because when your business knows everything about your business, When it's self-aware, so to speak, it becomes the ultimate teammate. So Rob has been working with these technologies for decades, and this is a concept that he is really excited about. It's a concept that we explore a lot in our book, Age of Invisible Machines. And so we brought in Elias Parker, who is the executive producer of that book and also of this podcast, just going to see where the conversation would lead. Hopefully you'll find this discussion as enlightening and as useful as we did. Before we jump into it, I did want to talk to you a little bit about some of the exciting episodes we have teed up for you in season four of Invisible Machines. We're going to be exploring the concept of organizational AGI more.

with Ben Gertzel, returning guest who you might remember helped popularize the term singularity. We're also going to be talking about organizational AGI and how it relates to organizational agility with Aaron DeSmet of McKinsey. And we also are going to be bringing you some more in-action episodes. Our AI Agents in Action episode from last season was actually our most popular episode to date. So we're really excited to be sharing more episodes in that format with you where you can see this technology in action.

And we're going to be digging into things like RAG systems and micro UIs. We've got a lot of exciting stuff in store for you there. For now, though, let's jump into this conversation about organizational AGI, or as you might soon come to enjoy calling it, Oagi. Invisible Machines is produced for UX Magazine in partnership with OneReach.ai. Named a leader by every major analyst group, OneReach.ai's generative Studio X platform was built specifically to facilitate the no-code orchestration of AI agents. GSX is trusted by global giants like PepsiCo, PwC, Deloitte, and DHL, but anyone can experience the level of agent orchestration we talk about on this podcast through interactive demos at OneReach.ai. Now, on with the show.

We're here today to talk about organizational AGI. You know, not to pat ourselves on the back here. This was a term that we came up with kind of out of necessity, right? Like we use the term hyperautomation a lot in the first edition of our book. And hyperautomation is quite a bit like organizational AGI, but as we just really dug into these ideas more for the revised second edition that we're working on, we needed kind of a better term. Organizational AGI, for various reasons, due to the conversations that we had with people like Ben Gertzel, who coined the term singularity, kind of led us to the space where we realized that organizational AGI, or as we sometimes call it, OWGI,

is sort of what organizations, it's something they can latch onto and focus on as they really try to make sense of all the AI going around the marketplace. Yeah, I think we just sort of waited for somebody to come up with something that we could use and it just hasn't surfaced. So we've decided, all right, well, we're going to use our term for now. And then somebody comes up with a better version or a more popular version. We'll just jump onto that. But for now, it's OHGI or organizational.

artificial general intelligence. And I think to set the stage, like, why do we need a term? You know, it's kind of annoying when people sort of jump on and create terms just for the sake of, you know, popularizing either the term or themselves. It's important to kind of understand why we think we need a term here. I go out to just, you know, what is general intelligence? Let's just take the artificial out for a second. It stems from this idea that humans really are remarkable at surviving in all kinds of environments. You know, the Arctic Circle, whether it's underwater, in the middle of the ocean, on an oil rigger, or...

in the Mojave Desert, as an animal, as a creature on this planet, from a biological standpoint, we're quite adaptive. And I think it would show pretty well if you looked at, you know, situations like a plane crashing on an island or something, and you sort of watch what we do, you know, from a macro level, we immediately start reorganizing the environment on the beach to suit our survival. So I think it's pretty clear that any organism's sort of meaning of life or goal is to survive and thrive. I don't think there's anybody in biology that would disagree with that. Might disagree with it as the primary directive, but I think we can all agree that it is, you know, it is absolutely a huge part of, you know, why we're intelligent and, and why that's been, and how that's been rewarded. And so you see that, you know, whether it's Cowboys Island or whatever, they're starting to reorganize.

their environment, they're moving sticks, they're building shelters, they're organizing it from whatever creature or creatures had sort of organized it for their survival and reorienting it for their own. And we could just say that that's a huge part of humanity is reorganizing our environment to make it more comfortable for us to survive and thrive. Whether that's on a molecular level, so we take rock and we turn it into AI, or whether we just move some sticks around and build a shelter. In essence, we're reorganizing this environment and we have a general intelligence that allows us to do this successfully. And so when we talk about machines having that general intelligence, we're basically saying that they would have the same level of intelligence as us so that they could survive anywhere on this planet. And it doesn't mean that their prime directive is to multiply, survive.

Um, that's, that's something we always ascribe as a package deal. Um, but, uh, I think those of us in the space understand that now the directive or objective can be to help humanity survive, uh, not themselves. And I think that's just an anthropomorphic, um, extension of, you know, our psychology, where when we start to anthropomorphize these things, we immediately start to ascribe our own. objectives onto them. And this is one of the things that I think we talk about a lot, Josh, at Hate Inc. We sort of hate and love anthropomorphization. Help me out here. We certainly don't love saying it all the time. Yeah. Anthropomorphization.

I think. Thank you, Ed. And that's what sort of led us to, okay, if that's AGI, if AGI is having the sort of general intelligence that we possess in machines to survive in any environment, to reorganize our environments for our own survival. And I think that's, you know, not getting down to the nitty gritty of, is that a collective intelligence or is that an individual intelligence? I think, you know, that's sort of another layer that for this purpose is not worth necessarily peeling and getting into, but I think we can understand that.

You know, we're not all capable of turning a rock into AI and probably no single individual is capable of taking a rock and turning it into AI by themselves, particularly on an island. But I think we can all agree that one way or another, general intelligence is, you know, our capacity to survive, whether it be as an individual or as a collective. We're a long way from that. I think there's too many in the space that think we've achieved that. I think some think we're closer than others. And I think...

I'll let you know what I think when we all agree on the definition of what it is. Yeah. Well, it's way too lofty a goal, right? For an act for an organization. It's too vast to really try and work toward. Yeah, and defined differently by too many people. So moving on from that, we say, okay, well, you know, it doesn't take that level of intelligence for, you know, a goldfish to survive in an aquarium or a plant to survive in a terrarium, right?

We can see that a level of intelligence to survive and thrive and be useful in that environment, in a specific environment, a limited environment, and one that is manicured, let's say, and managed by a human. requires a lot less intelligence, significantly less intelligence to survive. And so when we think about AGI, from a business standpoint, we sort of say it's a bit of a red herring. It's like, who cares if this thing, if this machine can survive anyone on Earth? Except maybe Google, who says they want to build an entity that could become a worker in any company. And any company being a company that does, you know, tours of Antarctica to building submarines.

then maybe perhaps for them, you know, the lofty goal of AGI makes sense. But for most organizations, what makes sense is to reach a level of intelligence with their machines and their group of machines or systems that's just good enough to really make a dent in their environment to help their company survive. So we look at this company as a subset, as an organism, as an entity, just like the law sees these corporations as their own entities. These entities' objectives are the same, to survive, really. That's ultimately their goal. They might say to deliver shareholder value, but you can't do that without surviving. They might say to grow.

But we see that growing is a form of surviving that's just thriving. So if we look at, you know, this world of companies and longevity, which is an interesting concept in itself, we sort of see that it's getting shorter. You know, as humans, I guess our collective intelligence is increasing. And so general intelligence as a group is the moving target. But the net effect is that our lifespans are increasing. But companies' lifespans are doing the opposite. Josh, you have some stats on this that I forgot, but maybe you can share. So yeah, in the 1920s, the average lifespan of an S&P 500 company was 67 years. By the 70s, it had dropped to about 30 to 35 years. And then currently it's somewhere in between like 15 to 20 years. So that's a pretty sharp decline. And obviously there are a lot of reasons that probably go into it.

You know, we talk in the book a lot about future-proofing your company, and this is sort of that same thing, right? Like part of being able to survive in a radically changing world and marketplace is to be adaptable and to have some level of agility. As we're kind of working on the second edition of the book and was looking into this idea of longevity. So yeah, I found a BBC article by Kim Gittleson from 2018, and she talked about the fact that Japan is home to more than, I think, 20,000 companies that are over 100 years old. There are even some that are over 1,000 years old.

You know, they have a word for these long-lived companies. It's Shinisei, I believe is how you pronounce it. Shinisei. Shinisei. But, you know, one of the things, there are a number of things that kind of contribute to their longevity that connected with ideas we presented in the book and things that we've talked about a lot. Some of them being like, these are companies that aren't solely tied to making profit. That's not their primary objective. And they also tend to be like very agile. And like she pointed to Nokia, who I think started out as a pulp manufacturing company and then got into electricity and then eventually mobile phones. Yeah, sort of breaks down a lot of the brand elasticity assumptions we have.

I think we're going to see that a lot. I think with the advent of technology, you know, we see that pulp manufacturers can become mobile phone makers. And I think with AI, it's the same thing. I think this is an opportunity that any company that sort of masters the self-driving company and the automation, you know, can easily pivot from whatever they're doing to being like the, you know, even top provider of AI tools. And I think a lot of them think they'll still be doing what they do, but actually success in this could lead to, yeah, the Nokia story, right? Which is reinventing.

in the company versus going out of business and being replaced. Yeah, agility is like a key piece of that. Elias, you shared a McKinsey report about organizational agility, which, oddly enough, has organizational AGI in it. That's the first three letters of agility. Agility also rhymes with stability, they pointed out. But what's interesting about these ideas...

of agility is that um you know a lot of what we talk about i guess when it started the conversation we were talking about hyper automation like being in this state where you're you're constantly finding new ways to automate processes within your company uh and that your workforce is heavily involved in that and it also and you're you're having to do that effectively you're having to start connect points within your company that currently aren't connected uh all sorts of like data and processes all have to be mapped out differently and you have to establish all these connections for this kind of thing to work but there's this ancillary benefit of of it kind of sets the stage for organizational agi because your company that's kind of the fertile soil, I guess, that AGI would grow out of, is that kind of culture and activity. And it's interesting, like, you can tie those, a few of the places that we've been in the discussion so far, you can tie them together, like, that the Nokia history is interesting for all sorts of reasons, but one of the more recent chapters, as a lot of listeners will probably know, is that they...

they sort of failed to adapt with smart guns. Like, I think there's a bit of like, they didn't go down with the ship, you know, they pivoted, they sold that portion of Nokia, they focused on the back end, they're still here. So the story's not over. I don't know that I would say that this is an example of failure or more like an example of success to adapt. Yeah, I think that particular example, like the history with Nokia with mobile phones in particular.

like that that piece of their company was a bit of a failure to adapt and then they ended up selling it off right but as a company like at a macro level they've been clearly they've been successful at adapting in general like because like to your point they're still here right but but of course like that you know in the same spirit as companies trying to advance on the laggy Right. Um, there's failure as a part of it and you have to like race towards that. Right. Yeah. I think that's the point where failure is going to happen. It's more about being really agile at a higher level to say, yes, like we want to be agile at mobile phones, but if it doesn't work out, we also want to make sure that, you know, that there's other ways to survive and thrive. And I think that's, that's intelligence, right? That's organizational intelligence. And I think that brings us back to organizational artificial general intelligence, which is can companies build this intelligence into their machines and their ecosystem?

And we think, yes, and we think that that level of intelligence is far less, far inferior to general artificial intelligence, but organizational artificial general intelligence, which is just general enough to thrive within that terrarium, that organization, your organization. And that's sort of the, like, you know, interim step one day maybe. you know, these companies will benefit from a general intelligence that can survive anywhere on the planet. And therefore these AIs can operate their company for them. But I think that systemic change comes with so much question of what is a company anymore. So, you know, to just sort of table all of that science fiction, you know.

I think at some point you look far enough out, it becomes like a docudrama, right? It's, you know, nobody really knows. I think for us, we want to stay practical and say, you know, organizational child intelligence is achievable. It's achievable today. It's something you can reach incrementally by leveraging AI today. And there's value, you know, as you build it, you appreciate the value of it. It's not a waterfall, build it.

and pray. It's along the way, you'll see benefits. You'll get that feedback. You'll automate things in intelligent ways. And we say intelligent, meaning ways that benefit from self-learning. And one of the things I think is really important as a measure of is your system, is your architecture, your AI architecture intelligent, is that it should be that with every innovation on AI, no matter who comes up with it, small company, large company, OpenAI, Google.

that your ecosystem immediately benefits and sees an improvement. So that it's not dependent on your innovation alone and that you don't have to build this stuff in, that it just immediately gets better. And we see this a lot with our customers where, you know, GPT 4.0 comes out and it just instantly improves latency, cost, accuracy for those things that we're using 3.5. And, you know, things just get smarter. All of the agents in the system immediately get an intelligent uplift and people start experiencing instantly its ability to perform improve. So as we sort of ride that AI wave, you could sort of think of your cognitive architecture as like the surfboard, you know, that allows you to ride that. And organizational AGI as the context, the ocean and things like that, the context that's needed to make sure that those agents operate within the framework of how you want things down within your organization, not

you know, how things are generally done in the world. I think that's a big part of OAGI is this idea that there may be a right way to do it. There may be a recommended way to do things, but that may not be the way you do things. And I think we're a ways away from trusting AI to just take over writing our employee handbooks for us, deciding all the rules on the fly and trusting that, you know, general decision-making and general rules of thumb are the way we'll run our companies. So we don't hand over that to the AI, which means we have to give these agents, AI agents, instruction on how we want things done.

And I think you could say the instructions of how we want things done and the information that's needed to do them our way is the operational general intelligence that you need to build. Whether that's through a digital twin, your organization, so it knows all the pieces and parts, all the past conversations and data, or whether that's the specific rules and processes that you wanted to follow. Please follow this process when people are going to book vacations, for example. What I started thinking about as you were explaining more of like what you mean by this term you've coined, right? Organizational AGI. What was going through my mind is like for builders and tinkerers, a lot of designers, people that will have their hands in creating AI agents and skills that are being sequenced by AI agents, et cetera, it'll be really easy to get distracted, lost, not even saying that as like a bad thing, like it's exciting. It'll be easy to dive into like all the possibilities and that might make it hard for a lot of business people to miss that like this isn't...

necessarily like a future casting we're nice to have sort of thing like there's actually like real necessity if you look at that that smp 500 curve like a very strong like um negative curve down indicating that the lifespan of companies is shortening dramatically and then you add ai to the mix like that was happening long before ai showed up in the business landscape that was already the trend now you're like throwing gasoline on that right yeah this is this isn't uh like oh wouldn't it be cool and we'd be more cutting edge in advance it seems like more of a survival thing than a um you know, cutting edge thing. Yeah. Yeah. I mean, there's a lot of questions as to why this is and whether this is good or bad. And I think that's out of scope. You know, if we all want to say this is, you know, maybe it's a good thing that we're evolving and turning over, like cells dying and new cells coming. Maybe this is an evolutionary advantage. And that's a good thing. I think from our perspective and the perspective of, you know, OAGI, I think we're basically saying, look, if you get into any of those companies, though, and you talk to the leadership and the people who are, you know, passionate about

where they work survival is what their objective is and so aligning with that whether it's good or bad doesn't matter to us the fact is we're trying to help them with their objective and their objective is to survive um so nobody's going to sacrifice their company for the greater good uh and and so i think it's you know it's great to have those discussions about whether it's good or bad that's fun but at the end of the day i think we're looking for a practical um term and a practical approach and process for uh increasing the longevity of your company just like you would you know humans and i think In an ideal world, we're going to see the introduction of AI. And we're also going to see the extension of life and longevity for companies start to go up. And I think one of the reasons is, you know, we understand that software is a form of automation, but it's a form of rigid automation. In other words, it takes a long time to build, deploy software.

and change it so we can see that in some ways we've you know benefited from software in a lot of ways we haven't because we now our business processes are sort of held hostage by the fact that to change them you know we need to change our software first and now we got to get IT involved and we did a project a while back with the Board of Education in Colorado and they're really radicalizing education they were sort of following a new process and trying to innovate education in the public system, which is really tough, right? And they brought us in to help them figure it out because one of the biggest challenges that they had was the software. And this new pattern that had been sort of proven out as a much better and more effective way to teach kids, one of the key ingredients to that was removing this concept of first, second, third, fourth, fifth grade and breaking things down into really small badges. You know, you would get a badge for learning how bees manage a hive. And so they would sort of break down

this standardized testing concept into littler pieces. Do you know this? Do you know this little piece? And then they would give more agency to the kids to decide within a realm of acceptability what they want to learn next. And one of the most exciting parts of this for me was that they also, by breaking it down into these pieces, were able to involve more of the community, take the pressure off the teachers from doing all the teaching to being more facilitators of education. And now a local beekeeper could teach the kids about bees. On a Saturday, the parents could take them there.

Um, he was aware of the curriculum that they needed to learn to get the badge. And then the teacher's job was just when the kids came back to test them, uh, to determine that like, did they, did they acquire the knowledge? So now you're like involving the community and parents into the education of kids. Um, and, and the teachers become more facilitators, but the key, the key problem here was every piece of software in the school started out with what, what grade is your child in? And like the software just didn't work.

outside of and to rebuild the software. There wasn't software out there that didn't say what grade is your child in. So to implement the system, they either had to like go back to paper and pens or they brought us in to kind of rebuild the system to be more agile. And so we can see that just in this example, like they couldn't reform their organization and they couldn't evolve it because of the software. The very thing that was supposed to make it easier made it rigid. And I think we can see this in organizations that our software is very fragile. It's very rigid. And it's preventing us.

from adapting um to overlap right like the the the decline of uh company longevity and the rise of software within companies there might even be some sort of connection there because you're right like software can speed up certain aspects of operations for companies but it really seems like it's become a very temporary gain like as as uh the environment shifts requirements change software becomes very much a problem. And so, like, as I'm looking at this McKinsey report, like, they've identified, it looks like a few different forms of organizational agility, and one of them is operational agility, which is, I think, sort of where, I guess, conversational AI can kind of step in and move companies in the direction we're talking about, right? Like, instead of having to speak to software in these very brittle, limiting ways, we can use conversational AI to kind of string the software together, and it can be the one to communicate and send information between these disparate solutions. And that sort of sets up a new...

You're always correcting me because I keep saying operational AGI instead of organizational AGI. And I think it's because sometimes... Yes, in that instance, I meant to say operational agility. OAGI works because it's... As part of organizational AGI. I think I'm going to keep that as just being interchangeable. Sometimes I do think operational AGI is sensible and organizational AGI makes sense because it is operational at the end of the day. Operational agility is what we're talking about. And what if software and AI can be the answer to that?

I think one of the keys to OAGI is also our AI software, AI agents' ability to operate our old software. I think that's like an interim step. If it's operating our old software for us, then it can start to, you know, over time, augment our old software and adapt outside of it so that we can move more freely. And so it may create features that... you know, that, that allow us to continue to use that old software without changing it. And then just slowly over time, the AI system relying less and less on the old software. So I give you example, let's say in Salesforce to manage my contacts list, right? And all my customer data, but I make sure that my AI agent is in between all requests to read, write, all the crowd requests, right? The create, read, update, delete.

Well, at some point over time, let's say it's logging all those requests, just like a person, you know, a person sitting in between. I say, hey, add Bob to the list. Okay, great. I could turn back an hour later and say, hey, is Bob in our database? And the person might just say, yeah, because they remember adding Bob, not because they went to Salesforce to see that Bob's there, right? So imagine that our AIs are just logging what's happening in the interim and eventually going, spending less and less time going back to those original data sources, needing to operate that original software.

less and less gradually over time where the end user never notices, that eventually doesn't depend on it at all and then get rid of the license and stop using it. But I think this is a big, big part that's exciting for me about this transition that AI operates our old software, but it also then eventually gets rid of it. I think, you know, people are worried about it taking our jobs unless, I think that's less likely than taking away our old software. I think that's... You know, that's something we all probably in most cases would be pretty happy about. Yeah, it can become the software and it can actually be truly soft and pliable. Exactly. And I think companies with OAGI, if they focus not on replacing humans, but replacing that old software, they'll find that there's humans and the company will get behind that. You know, there's a lot of alignment there. There's a lot of cost savings to be had and a lot of benefit to those people because it'll make it easier to do their job. So, OAGI is not, in my mind, you know, an objective to reduce your...

your labor costs as much as it's to reduce your software costs and grow and optimize. if you use conversational AI correctly to kind of give yourself operational agility, then you can actually start to move into these other areas that McKinsey talks about, like portfolio agility and strategic agility, where, you know, if you think about something like, I guess the Nokia shift from pulp to electronics, like if you're, if you have operational agility in place within your organization and the market does a 180 and you need to completely shift focus, it's going to become infinitely easier than if you're tethered to all this like legacy software that'll just encumber you. Some of the other attributes of these companies that have been around for hundreds of years in Japan is that they often tend to be small and family run.

Which is interesting. Not to say that our book advocates for all family businesses, but I think there is something valuable to businesses acting more like family businesses. I think it ties into a strategy that we've outlined for kind of reaching a state of... where you're leaning heavily on your workforce to help you get there. Like these are the people who really understand the ways that these kinds of technologies can move an organization forward. And by creating automations and giving them tools to make their jobs better, you're giving them a better work environment. You're showing some level of care at least for them. And so then it might move more into that realm of like the family business where, you know, you're invested in the success of the business because you feel like you have a stake in it. And there are clear indicators that its success becomes your success. And I think that relationship seems to be absent in a lot of the companies that you...

that are dominating different industries today. So that could be another kind of interesting elements of this that OAGI could somehow bring about. Yeah. Yeah. I almost like when I think about companies today, one, you know, analogy that comes to mind is like, you know, sitting in a bobcat, you know, with that cage around you and you're like, yeah, I mean, I can move stuff. I can dig faster than a human. Like, look at me go.

completely unfair, you know, a person in a Bobcat against a person with a shovel, right? The Bobcat beats the shovel for sure until the job changes and it's not about digging anymore and it now becomes, you know, about building something or, you know, they hit a water pipe, right? Now this thing's useless and I think we can sort of think of this Bobcat as our old software, the software we've... We've bought and built around our business with the idea that the way that we succeed in our business will be fairly static long enough that we'll see an ROI from the investment in the software. Explore, exploit. Yeah, yeah. If the business becomes, you know, and the business landscape becomes so agile and dynamic, you get stuck in that old software. It's sort of being like, you know, stuck in a cage as it goes underwater, you know, you just won't be able to get out.

And I think that's, that's funny. Like our old software was our tools to gain an advantage. I think now they're like a gigantic boat anchor. And that's, that's why I'm always so curious. The, you know, question between large companies that have capital to adopt AI and invest in AI against the fact that they've also got these boat anchors attached to them, which is their old software. Like the small companies are more agile generally. Yes. Agile almost because they, they, they don't have the software. Like they can, they can choose it.

Yeah. So if they choose AI, then yeah, they can, but then they don't have the capital. And so we see, you know, a big question, you know, of why I don't like to forecast the future is to say, you know, I don't know, does capital beat out agility in this case? You know, I guess somebody might argue a small company with capital. Maybe that's the foe, right, to be wary of. I like the nautical theme too, right?

in terms of a boat anchor, like the point solution approach that people that a lot of companies are taking with the technologies associated with AI is almost like just barnacles on that anchor, right? Like this isn't about decorating your anchor. This should be about getting out from underneath all of the software that's weighing your company down. Like this is a golden opportunity to rewrite your future in that way. Yeah. And the key that I think people think like, oh, get rid of it. You know, at least that's what people usually say when I say that, what you just said. And no, it's put AI in front of operating it. Have AI operate your old software. Create new UIs that are conversational.

between your employees, your customers, and those systems. Abstract them. And then eventually they'll disappear. And I think that's hard for companies because from their perspective, they've got to pay for their old software. And now they've also, in addition, got to pay for this new AI software, this abstraction. I see that being very, very hard. And that's the counterpoint to young companies losing to big ones because they have the capital, but if they're not willing to deploy it in ways that increase the cost of their software in the short term so they can decrease it in the long term, if they're not willing to use it to do that,

then they will lose. And I think that's a really big challenge. They sort of say, how do I apply AI without increasing the cost of operating it? Because I am stuck using Salesforce and paying my Salesforce license for the next few years. And it is those that make that investment and say, look, you're going to have to keep paying that and you're going to have to pay this new tax, which is AI. But over time, that's going to pay off.

anyways pound foolish thing to some of that yeah or short-term long-term yeah like it's not it's not gonna it's not gonna get to the savings for the next quarter um yeah i think that's like so many people seem to both even in like personal use and organizational use are looking at these technologies as some sort of shortcut yeah um but it's really it's something else that it does often require those uncomfortable doubling your investment conversations and actions at least temporarily i think no matter what we're all you know we're all going to see a moment we've got to reconcile this concept of longevity of a company um You know, we've got to decide, or each company's going to have to decide if longevity is truly their objective or if short-term, you know, stock improvements are what they're all about. They're going to abandon ship the second they can. And I don't think that gets solved, you know, by the leadership alone. I think it gets solved by the investment community as a whole that decides that they don't want to, you know, be constantly shuffling their money around from one asset to another. And they're like, like this idea of having these long-term assets that they can rely on.

So as long as they reward that short-term thinking, as long as the market's rewarded, I think it's going to be a little tough. But there are companies we see like Palantir that basically, Amazon was a good example too. They just say, look, we're going to lose money for a while. We're going to invest in this. It's not going to look good. But in the long run, we're going to win. Whether it's free shipping on the Amazon side or Palantir just implementing like five to 10 big customers and making sure that they get that right. And they do it their way by paying for it themselves. So I think we'll see, that's the mentality that's required.

And I think what they're ultimately doing is creating some level of intelligence. If it doesn't result in organizational AGI, then you've just invested in random acts of AI. And I think that's what everybody's worried about is that they're going to make these investments in AI all over the company. But these are just bolt-ons to existing products. And they're just two letters that have been added to all of these product names. But at the end of the day, they've just spent a lot of money on random acts of AI building that has no strategy or unification or ultimate goal.

and doesn't benefit from this idea that AGI improves and that their whole ecosystem improves along with it. That's ultimately so much more costly, right? Like you mentioned the Amazon analogy, once a company gets it right, it won't matter so much how much money they spend getting there because they will have probably come to quickly dominate their entire industry. And even other industries, I think, will be forced to take notice of how radical that shift will be once someone really gets it right in the way that Amazon did. they took they took the risk and made sacrifices and then once the payoff finally came like it changed so much of the world as we know it it sort of requires somebody taking the reins and almost like a centralized effort to implement a decentralized approach right like like i recently heard from a bunch of startups that were in like a gross mode like they're they're growing out of like being what a lot of people think of as a startup that are that realize that they're using like uh asana and click up and monday.com like they're paying for all of those and like didn't know it right yeah

it's like this random, like they were just going, people were just trying to get shit done and it wasn't like a coordinated approach. And that's exactly what's happening with a lot of AI, whatever it's right now. We look at one of our customers, which unfortunately won't let me name them, but the really big beverage company. And they just began simply with their own IT team, you know, and they started with some of the executives that were behind this and building things that made their lives easier. And then just slowly, you know, building a culture around it. I think the rest of the team saw the executives using it and they were like, well, we want to use it. So it was sort of modeling the behavior instead of telling the rest of the company to build and use this. They were using it. Demonstrating. They built an AI for them to use, named it, and had to do things like organize meetings and things like that for them. And they started small and they sort of modeled the behavior that they wanted the rest of the technology team to do. And then more recently, now we're doing lessons.

and classes to hundreds of employees as they're learning to build skills and contribute and understand how the system works. So I do think it's about, I guess it's like anything, you build the model car in your mind, you build the vision, but it doesn't mean that you built a plan to get straight there. You just make sure that things you're doing, you know, make sense that they're on course there. Yeah, in service of that. And I think that's, maybe that's where people get off course. I think there's a big danger in...

I think it's probably the biggest danger right now for large companies is bolt on AI. Yes, it can elevate the value of your existing software, but if you do it with this idea that that is a strategy and you let your vendors decide your AI strategy for you and you just depend on them, how are you going to end up in a place where you don't depend on them? If that's the goal to not depend on your software vendors and your entire AI strategies depend on your software vendors. Yeah, you're going to be in trouble. You have to check the box. We're doing AI. We have ChatGPD. What about employees? We've identified that it's critical that employees take part in this journey, but why should they care? Because they might see organizational AGI as even a bigger threat than generative AI. If an organization is suddenly so good at all aspects of operations, why does it need me? And if that's the endgame, why should I take part? Yeah, why do they care? If you like your job and...

You know, what we can also always say is what's inevitable is change. Embracing it allows you to, you know, reshape your new job and make it something you like also. I think this is like a big reason about, a big reason why I got into AI way back when is I knew it was going to reshape everything around work and life. And I didn't want to, I don't want to be subject to it. I didn't want to be the object of the change. I wanted to have a say-so.

in how this change happened. I wanted to know enough about it that I could navigate myself. And I think we see that a lot now. We see a lot of people understanding that it's not AI that's going to take your job. It's a person that knows how to leverage AI better than you that's going to take your job. And I think it's very true. And I think that's probably the biggest reason for any individual to do it is to understand that change is inevitable. That you can't put generative AI back in the box. Slowing it down is just ripping off the band-aids more slowly, which is...

you know, great, but probably unlikely. Well, you mentioned survival of the fittest when you were talking about kind of the way that we view capitalism. And we've kind of talked about this before, right? Like maybe not on the podcast, but that Darwinism is kind of misrepresented in that sense, because I think when we think of survival of the fittest, especially in terms of capitalism, we assume it means like the biggest, strongest, strongest, the most dominant thing that could just beat you into the ground is the thing that wins. But really, it's more about the thing that's most keenly adapted to its environment. And we're entering a phase now where the environment itself.

is going to be constantly in flux. So it gets back to that organizational agility, which is, of course, like a big part of Awagi, where it's like the fittest means the most nimble and the most agile, capable of changing shape and, you know, adjusting to a shifting landscape. Yeah. Just because I have a habit of, as people could tell, going on and on and on. I said the Darwin thing, but really wanted to caveat that to say, most people misunderstand that as like Darwin saying it's the one that can dominate its environment when fittest is more meant to be about the puzzle piece fitting in.

to your environment, not dominating it, not owning it, but fitting into it. And so that's a right sizing versus a big sizing. The best fit, not fittest. Yes. The long proboscis that can get down to the bottom of the tall flower. Yeah, it's not the one that just dominates all of the things around it.

In fact, we can see that with dictatorships, that's not necessarily the best way to survive and have the longest life. So yeah, I resisted that clarification, but thank you. I think it's important to understand that operational and organizational AGI is about fitting into the marketplace. It's about finding market fit, not dominating every market. But it's finding the places that you can fit nicely into. We hear about that market fit as a concept for startups. Different kind of fitness. Right. And I think that they're getting more to it. They didn't call it market domination. They call it market fit. Can you fit into that market? And when you do fit, it actually fits pretty nicely. McKinsey's take on things. It's about agility. You could be fit in a physical sense. You could be incredible at hurdles or incredible at this.

athletic endeavor that athletic endeavor but to have agility and be sort of this like adaptive multifaceted athleticism has more to do with agility and not necessarily specialization i don't know if they connect the dots this clearly in their in their work but the idea of achieving operational agility would seem to lead to portfolio agility that they talk about strategic agility where you're now able to not just adjust your internal operations but also move around resources and and rethink the way that you fit into this kind of amorphous marketplace that we're surrounded by at the moment. Yeah. One of the brain twisters that I love the most about this is, you know, so many times I've been on software projects and I'll speak from the software lens, but so many times I've been on software projects where developers are building something and they don't even know what the objective is. Like, I can't tell you how common that is. And I guess most people would agree that are in it. They're all nodding right now. Like, yes, that happens all the time.

But with this new generative AI approach where you're not necessarily giving instructions to the machine, but you're giving objectives to the machine and then just providing guardrails, you know, so that they have to meet this objective within these guardrails. It's sort of like playing the game of Go or anything like that. You create the limitations. You can't go outside off the board, but you have to get this thing accomplished. If you're now the new programmer, right? The new programmer mindset, you can't start without knowing the objective because your job is to communicate the objective as clearly as you can to the machine.

And any ambiguity in your ability to articulate the objective will result in the system not operating correctly. So this will be really fun because as much as, you know, folks have really tried to involve developers and developers, I think I've always wanted to be in those meetings and in those conversations about what the objective sure shouldn't be. Now it'll be necessary. And I think that's, Josh, you know, to your point, you kind of started to keep bringing up the idea like... You know, is that a new way of IT departments and development departments operating? Or is it the end of the IT as software department? Just like we saw the end of the computer department. Everyone has a computer. Everyone's in the computer department. In essence, no one's in the computer department. Yeah, we're all in the IT department. We're all writing software all the time because we're having conversations with machines. We're giving them objectives. They're writing ephemeral programs. Yeah. And so it's this interesting world of, is it that...

IT will change or is it that we'll all become developers or, you know, likely both for a while. So, I guess that just leads to the fact that operational AGI requires, you know, overall organizational cooperation towards it. It's sort of, at some point, you need, you know, you don't have to start this way, but at some point you need buy-in across the board as creating operational AGI. It becomes a critical objective of the company. Everybody's got to get on that page. All right. Well, thanks for hanging out with us again here on Invisible Machines. Don't forget to like and subscribe wherever you get your podcasts. Always remember that you can watch this podcast, too, on our Invisible Machines YouTube channel. Head over to uxmag.com. We have a podcast page there where you can explore our back catalog and you can look at episodes centered around popular topics. Many, many thanks to our incredible production team here in the U.S. and in Ukraine. Elias Parker, Michael Lifinoff, Allison Harshberger, Alis Lesarenko, Vera Pakhodkov, and Darina Moskovchuk. Till next week.

Frequently asked questions

How long is this episode of Invisible Machines podcast by UX Magazine?

This episode is 1 hour and 0 minutes long.

When was this Invisible Machines podcast by UX Magazine episode published?

This episode was published on July 25, 2024.

Is there a transcript available for this episode?

Yes, a full transcript is available for this episode. You can read the complete transcript on the episode page.

Can I download this Invisible Machines podcast by UX Magazine episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.