AI Requires Humans and Guardrails

Bryan Cassady (00:01.214)
That’s better than my answers.

Patrick Spencer (00:02.638)
All right, if you’re ready, I’m ready. So, hey, everyone, welcome back to another Kitecast episode. I’m Patrick Spencer, your host for today’s show. Joining me is Brian Cassie. Brian, this is going to be a great conversation. Thanks for making time today.

Bryan Cassady (00:08.796)
I’m ready to go.

Bryan Cassady (00:21.746)
Thank you very much. I’m glad to be here.

Patrick Spencer (00:24.346)
He’s actually across the pond, I think right now over in Europe. He is the director of Global Entrepreneurship Alliance and is on a mission. If you look at his LinkedIn profile, you’ll see this. The train, is it, Brian? One million entrepreneurs on AI by 2027.

Bryan Cassady (00:43.39)
That’s looking a little close right now. Yes, that’s the objective, but I’m at about 90,000. So next year is going to have to be 900,000.

Patrick Spencer (00:51.245)
90,000 is pretty good. Well, to our listeners, we’ll obviously let you know how to get in touch with him at the end of today’s podcast. Get in touch with him, help him get to that million mark. He has an impressive background. He’s built eight companies across six different countries. He’s worked in with over 30,000 leaders worldwide on a variety of different business initiatives. He’s the author of award-winning books, one called Cycles. and the generative organization. We’ll talk a bit about those today and we’ll also include links. If you’re on YouTube or Apple or Spotify, you’ll be able to go to the bottom of the blurb and click on those links so you can check those out on Amazon. He’s the creator of something called AI for Innovation Toolkit for Human AI Collaboration. I’m going to ask him to explain to us what that’s all about. He’s a professor, keynote speaker at Chicago, Bentley, a bunch of other places. KU Leuven and he’s led accelerator programs at European Innovation Academy and the Founder Institute. has MBAs, not one, but two, Cornell Johnson, Varalec Business School, and he also worked on a PhD. I forgot if you finished it. Probably got sidetracked with all the other initiatives you work on, Brian.

Bryan Cassady (02:11.1)
No, shoot, I didn’t finish it, but know, life goes on.

Patrick Spencer (02:17.678)
Very true.

Bryan Cassady (02:17.918)
But the person would be happy if I finished my PhD would be my mom. She wants to have a doctor in the family. But the good news, I’m actually on the stage in about a month with a Nobel Prize winner. And my mom is more chuffed than anybody.

Patrick Spencer (02:23.79)
Mmm.

Patrick Spencer (02:31.256)
wow.

Patrick Spencer (02:34.892)
Yeah, she should be. So what’s the topic about?

Bryan Cassady (02:38.162)
I’m talking about when help comes too early. And the idea is talk around, we don’t think, but we ask AI for help. And if we don’t think first, we get really bad answers. And I was sort of a long shot. I didn’t think I would get into the conference. There’s about 2000 applicants for 10 spots. And I did get in. So I’m excited about that. And my big cheerleader when I do something academic is my mom.

Patrick Spencer (03:01.57)
Yeah.

Patrick Spencer (03:06.922)
I think mine is still as well. We’re, blessed to have good parents, aren’t we? makes, makes a big difference. So that’s really interesting. I would say we also get lazy if we’re not pushing AI at the same time. Right. I’ve found that those who, you know, push AI and constantly want to learn more and are reading the outputs, you know, they, they’re learning something and they’re pushing the coin versus just, as you said,

Bryan Cassady (03:11.932)
Yeah,

Patrick Spencer (03:35.694)
help is important from an AI perspective. So that’s maybe a great segue into our conversation. I have this concept you write about in some of your writing, the 1 % problem, why ideas don’t translate into action, I suspect is a bit related to the topic we just mentioned.

Bryan Cassady (03:55.034)
Jesus. So this happened about a month ago. I was out at a cafe with a friend of mine and he said, Brian, tell me about your book. I said, things are going great. I’m on track to, I’m a few bestseller lists, got some good reviews. said, how many will you sell? Well, I said, well, you know, a bestseller isn’t that many books, maybe 15, 20,000 copies. And then he grabbed a piece of paper. And he wrote a bunch of numbers down. said, how many people do you think actually are going to use your book? And he gave me the answer, which is 150.

Patrick Spencer (04:31.1)
wow.

Bryan Cassady (04:31.998)
And I said, what? That is impossible. So he went through a process and how many people are going to read it. And I said, give me your number, Brian. I said, ah, 25%. He said, you’re optimistic. Average business books are read about 10 % of the time. How many, what percentage of understand it? Oh, I’d say 50%. I write pretty well. All right. So we keep on going down. How many people actually plan to use it? Another 50%. How many people actually use it? 50%.

Patrick Spencer (04:34.306)
What was his algorithm?

Bryan Cassady (05:02.214)
And pretty soon you get down to the 1 % problem, which is, you, you spend a year of your life writing a book and the number of people actually use your book is less than you put in an auditorium. And that got me thinking. And about a week later, being a little bit of an entrepreneur, said, look, I’m going to do this. I’m going to give my books away for free. And what I’m going to do is to make it easier for them to use it.

Patrick Spencer (05:05.55)
Hmm.

Bryan Cassady (05:31.6)
I’m going to bundle my books with AI tools so that in fact, not only do I get more books in people’s hands, but I get it make it easier for people to use their books. And what we’re seeing right now is I get about a thousand downloads per week, which is good. And the usage rate is around 20%, which is fabulous. So, but you know, this is something that comes up everywhere you go. It’s not just with books, it’s keynotes.

Patrick Spencer (05:46.232)
Hmm. Yeah.

Bryan Cassady (06:01.828)
It’s trainings, it’s internal sessions that you run in your company. And there’s a huge gap between intention and action. And I think the rule or the new rule of AI is going to be to help close that gap. Because, you know, if you know something, it’s hard to apply it in the old way. Just try it. Tell me the last book you read. And tell me about a framework which is in that book. Ooh, now tell me about a book you read two years ago. So, seriously, I can’t remember anything anymore. But if you say, look, I remember this book and it sort of about this, and you ask AI, you know, help me apply this framework, you can actually take those ideas that you had and put them in action really quick. And I think, I think just going to change everything around. I got a lot of hate mail from fellow authors.

Patrick Spencer (06:37.73)
Yeah.

Bryan Cassady (06:59.742)
saying, what are you doing? I said, well, you know, I’m trying to promote my business. My business is to help companies do stuff. said, but you know, you’re killing the book market. I said, but no, it’s not killing the book market. This is actually giving your book new life. Your book actually can now be used. Isn’t that great? And maybe 50 % say yeah, and 50 % say, no, I don’t like what you’re doing, Mr. Cassidy.

Patrick Spencer (07:27.256)
Hm. Hm.

Bryan Cassady (07:28.734)
So, you I think it’s an interesting question, but, you know, really we have to move from a little bit of our vanity metrics, which is how many books we sold, to how many people actually use the things that we share. And frankly, that’s a bit humbling.

Patrick Spencer (07:48.974)
That’s interesting for our audience. Many of them are CISOs and so forth, but they consume valuable content. And I won’t name names, but I The different companies and organizations that publish regular annual reports, guess what? Most of them are still gated. We stopped gating ours here at Kiteworks for whatever it’s worth about four and a half years ago when Tim and I first got here. He said, we’re going to stop gating this stuff because we actually want people to read it. Because when you gate it, then

Bryan Cassady (08:01.822)
You

Patrick Spencer (08:18.976)
A lot of the audience will disengage and I suspect the same is true from the book market. You have to go buy the book and if you give it away and then this concept of, you know, taking the idea and having AI tell you someone should do this with today’s podcast, I guess, probably tell you what the key takeaways are so you can operationalize it in your life or in your businesses.

Bryan Cassady (08:41.884)
No, I think like you listen to this podcast, take the transcript and say, look, give me three things that I can do tomorrow based on it. Isn’t that amazing? By the way, there’s a interesting silver lining of giving my books away for free. I’m selling more than I’ve ever sold before. So actually what happens is you give away the PDF, people actually want to buy the physical book.

Patrick Spencer (09:00.652)
Hmm. Hmm.

Patrick Spencer (09:07.35)
Interesting.

Bryan Cassady (09:08.743)
So not only do I get more people to use my book, but I sell more books.

Patrick Spencer (09:14.188)
Win-win.

Bryan Cassady (09:14.589)
Yeah, that’s a deal. It’s a deal.

Patrick Spencer (09:17.976)
All right, the link to the podcast is at the bottom. So click on that and you can get it for free to Brian’s point, but you also can buy a copy on Amazon.

Bryan Cassady (09:26.686)
Yeah, I’ll give you the link in the podcast. But I’m curious, do you measure how many people read your reports?

Patrick Spencer (09:36.36)
we do. And the numbers are certainly more than they were when they were gated. so you you send it out via press release, you put it on social, you blog about it, you put a sub stack or two out on it, maybe more, you pitch it to the media. and if it’s a darn gated report, then fewer people are obviously going to open it and register to actually get it. And then, half the time you end up with Charlie Brown or Lucy, Lucy, submissions that are pretty useless at end of the day. So since we stopped getting that, know they get engagement and we get much better media pickup. There’s a lot more people talking about on social. suspect you’re seeing the same thing with your books.

Bryan Cassady (10:19.09)
Yeah, same type of thing. And I think in the end, what it does is it makes things a bit transparent. So in fact, if your report is delivering value, the business will come to you. If my book delivers value, the business will come to me. If it doesn’t deliver value and you take that gate away, then you’re really in trouble.

Patrick Spencer (10:40.216)
Well, let’s drill down into how businesses, because we have a bunch of listeners on our podcast today, I’m sure are business leaders and they’re struggling with AI adoption within their organization. Every day you feel farther and farther behind or you can be. You have this quote in one of your books, you talk about AI must earn its keep and you call out that eight out of 10 organizations aren’t getting real results from AI. And your claim is they’re chasing the tools without the strategy or the change management that comes with that. So can you describe the problem and then what’s your solution?

Bryan Cassady (11:23.762)
So I think the problem is bright, shiny lights. It’s like, wow, look at this tool. And this fear of missing out, and you try this, and you try this, and you try that. And what happens is we end up using tools before we have objectives. So in fact, that’s the biggest issue. But then the second issue comes up is when you’re swimming in a pool which is one inch deep and a long wide, you don’t actually get much out of the one inch, which you should be doing. is using one thing and learning to use it well. And what I see, the real opportunity comes up is when people go from objectives backwards rather than tools forward. So you say, look, I want to achieve this. OK, now what do I need to do to get there? And then you actually get results. But too many times, it’s an external consultant, maybe somebody like me, who came in and said, wow, this is an amazing tool. And you go install the tool, and nobody uses it. Or they get confused. Or worse yet, they actually end up working harder when they have AI in the bunch. Because in fact, you can generate so much content. I don’t know about you, but I can get a 50 page report from somebody. Are you actually going to read the 50 pages?

Bryan Cassady (12:50.722)
And so I think there’s a question of earning its keep. And I think the opportunity for the future is to stop using AI and start working with AI. You start thinking about it as a means to an end rather than an end in itself. And then you see different results coming in.

Patrick Spencer (13:14.062)
So how can organizations, you have all these tools, should they measure the usage of their tools? We have a bunch of AI tools in our environment and we can measure the use of some better than others. The outputs obviously is how business should measure the results at end of the day, but just knowing who’s using it and how are they using it? Are organizations mature enough to do that today?

Bryan Cassady (13:40.872)
Most are not, sadly. I think the real metric is not usage levels. It’s not efficiency levels. It’s effectiveness levels. And usage is relatively easy to measure. Efficiency is sort of easy to measure, but effectiveness is hard to measure. And the real measure today should not be, can I write a report six times faster, but can I write a report that is more convincing? more useful and more valuable.

Patrick Spencer (14:17.816)
Very true, very true. So you also talk in terms of, I have some notes written down here, the integration portion of AI. When you pick these different AI tools, they kind of be integrated into your business process. Because if you can throw a bunch of AI tools to your point out there, but unless they actually are integrated into what matters to your business and your workflows and so forth, then they don’t really deliver true value. That’s the first part of the question. The second part of the question is, you know, with AI agentics, are we beginning to see that come to fruition?

Bryan Cassady (14:53.958)
Yeah. So, you the integration for me is about something that’s kind of non-sexy, but I still continue to talk about it because I think it’s really important. Every business is a system. It’s a bunch of things working together. And what you need to do is you need to think about how things integrate and how they work with one another. If you you can pump out more material faster and you’ve got a bottleneck somewhere down the lower, it doesn’t help. So An integrated company is actually a company that looks at the whole system and identifies where are the weak points in the system and improves those weak points. And sadly, so few companies are interested in systems thinking because it requires hard thinking. But sometimes hard thinking is the right thinking.

Patrick Spencer (15:46.958)
What’s your perspective? And I this goes into this fact that it just can’t be AI. You got to have a human component. They go hand in hand. That’s one of the elements that you argue is required in order for you to have a successful AI strategy in your books. And we’ve seen a lot of discussion about AI agents in the last couple of weeks, and we’ve seen this talk. the software as a service providers, they’ve really struggled. Some of them have lost 60, 70 % of their value in a matter of less than a month. What’s your opinion in terms of AI agents? And is it a threat to these SaaS companies like the market has initially perceived or as things gonna calm down, they’re gonna actually realize that AI agents, when you use them in concert with people actually, is going to enhance productivity and make these SaaS solutions more useful and effective at the end of the day.

Bryan Cassady (16:49.182)
So actually there’s two questions I think you’re asking there, Patrick. know, one, is AI gonna kill, you know, the software as a service market and things like that? Yeah, it is gonna kill it because you can build stuff so quick, so easy, and you can customize it for what you’re doing. I think what’s gonna happen is there’s gonna be a whole new market of modular software where in fact you build this bit and they sell you a little module and you build another bit. And I think that’s harder to get money out of than other things. The other question is, what is the future of agents? And the future there is only as good as the direction we give them. We were talking before about rockets on the back of bicycles. You send out an agent without poor instructions. It can knock things down. It can break things. And that’s not going to help your business. And where it comes back to is if you want these things to work well, you have to come back to some business fundamentals, which is figure out what your strategy is. What do you want to do? Why do you want to do it before you let the tools go out and do it? And I don’t know if you’d agree with that, but I think it’s, you know, what is old is new again. know, strategy first, my friend.

Patrick Spencer (18:10.85)
you

Patrick Spencer (18:16.75)
So you bring up an interesting point. I’ve been thinking about this. I don’t know I have an opinion or I certainly don’t have answers on the subject yet, but you with these, you you mentioned that AI is going to kill these SaaS models. You know, what’s going to happen to these companies? What do they need to do to pivot to be successful in the marketplace in your opinion to ride this AI agent wave?

Bryan Cassady (18:39.934)
I think what’s going to happen is they’re going to all be AI agent companies pretty soon. I mean, it’s just marketing. mean, most of the stock market valuation is not the functional aspects of the organization, but what is the marketing around it? So they all become AI augmented, AI agent this and AI agent that. But the fundamental thing that’s going to come back and determine long-term value is do they add real value in the process? And I think what you’re going to have is a few mega softwares that add real value and integrate very well. And I think a lot of the small stuff is just going to kill them along the way. So it’s going to be a few basic systems, which you use some customized systems and that sort of middle ground of, you know, 10,000 users and stuff like that. think that they’re going to have a tough life going forward.

Patrick Spencer (19:32.142)
And the market for modular solutions, these companies may go away, will actual individual businesses be the ones to build it themselves? Or will there be a market of organizations that are developing these modular pieces that can be integrated into these larger software deployments, in your opinion?

Bryan Cassady (19:55.71)
God, I wish I knew because then I could invest right. I think what’s going to happen is there’s going to be a lot of modular stuff that comes out. it’s going to be how well you play with others is going to be a real determinant of the future. And do you make your module open-ended? And the companies that are trying to build something that does everything from A to Z, I think they’re going to get killed off as well.

Patrick Spencer (20:23.278)
So it’s important to focus on specific business issues and things you’re good at rather than trying to boil the ocean.

Bryan Cassady (20:31.358)
Yeah. I mean, the thing is, you know, you know, if he came back in five years, everything they’ll say will be wrong or part of it will be wrong. But, you know, that’s the hard thing about the future is to predict the future. But I, I sincerely believe that there’s going to be a movement back to value and people are going to start saying, look, it’s not because it’s agentic, it’s good. It’s good if it delivers value. And I don’t care if it’s agentic or something else.

Patrick Spencer (20:58.254)
How do you determine, you we have a lot of agentic projects and I’ve worked on some myself and some we’ve mothballed and said that was a bad idea or that’s not possible or that won’t work now, it might work in two years. How do you, you know, prioritize those projects? One, and then how do you decide when to pull the plug on one that just isn’t going where it should be going?

Bryan Cassady (21:26.782)
I’m going to pass on that one because I don’t know. I think it’s a philosophical question. I think the thing is, if you start with a really, really clear objective and you can see that you’re making progress towards that objective, I think there’s a good reason to continue. And I think most innovation and most product development is incremental movement in the right direction. And I think the point where you have to kill it

Patrick Spencer (21:27.672)
We’re too philosophical.

Bryan Cassady (21:53.53)
is when you say, look, we’ve just been playing around here, and it’s just causing more work and more wasted time, and we’re not getting any closer to our objective. And that means having the rigor and the discipline to define objectives upfront and have real kill points in the process.

Patrick Spencer (22:11.054)
Yeah. Hmm. Those are good suggestions, despite what you initially said. Yeah, I think you could answer it. So you we’ve talked a bit about the importance of humans plus AI, which you argue in your book. You know, I think there’s a quote you have in one of them where AI didn’t kill the business. You know, it changed its job. You know, what do you mean with that?

Bryan Cassady (22:35.454)
So the quote was there is AI didn’t kill the book. It changed the job of the book. And I think in that specific case, AI is a way to augment some information. But I think a really important thing to remember is it’s not AI or humans or whatever. In fact, the value, it’s like a relay race. And you have to decide what’s going to get you to your destination. Should AI run first? and then hand off to humans. Should humans run first, hand off to AI? Or should it be all AI, or should it be all humans? And the value here is about the handoff and deciding the order in which you do things. And too many times right now, it’s all AI first. And I think that’s the most silly, silly, silly advice. know, people say, we’re an AI first company. I said, God, please be a thinking first company. and then be an AI company. So it’s a question of order and the degree to which you do things in the right way. I actually have a talk coming up where we’re talking about AI or humans. It’s not the question. In fact, the question is how do you integrate the two? And one of the things I’ll talk about is chess, for example. Do you know who wins right now in chess, AI or humans?

Patrick Spencer (23:58.798)
Hmm.

Patrick Spencer (24:05.934)
Humans got to be because the humans are guiding the AI. I assume that’s going to be your answer. No, wrong answer.

Bryan Cassady (24:09.886)
So it’s still the AI that beats the humans in chess. Okay, AI plus a human, a chess master who wins. It’s always the AI plus the human. And in fact, even in the areas where humans are less strong, oftentimes they add a unique skill or a new competence that AI doesn’t have. You know, it’s the question about how do you put the two together to work effectively? the question there is not one or the other, but how do you put them together in a smart way? How are you holistic? How do you look at your system and make the system that delivers the results that you want to get? God, I feel like I’m on my soap box here. God. But I believe… You know, what I see is I go into companies and I see them so pumped up and so excited about this and said, my God, what are we going to do about AI? And I said, actually, what are you going to do about management of your organization and then worry about AI? Get the fundamentals right. And I feel each time there’s a fad that comes along, we sometimes forget what’s important. A few years back, I ghost read a book called Flavor of the Month. And Flavor of the Month was anything that somebody was reading about right now and how that becomes the new strategy. And really what there is is there’s old flavors which come back again and again. Clear thinking, effective management of people, good team management, integration, diversity, all those things are more important than ever before.

Patrick Spencer (26:07.534)
Great point. Now I’ve seen this, you’ve seen this where just doing AI for the sake of doing AI or doing something that automates some process, because you could just for the sake of automating it, organizations can dive off that board and chase those rabbits forever. There’s some instances where it probably decreases the productivity of the person, I suspect, or the org. And in others, it just makes no impact at all. How do organizations decide where to apply AI and where to say, yeah, we could do it. It’s quite possible, but it’s a waste of time or the benefit, the amount of time required to do this outweighs the benefits we would see.

Bryan Cassady (27:01.246)
not going to give any names, but I was doing an event with a gentleman. And we have an event which has about 100 participants coming to it. And he spent about a week automating all the inboxes and all the outgoing mails and things like that. And he did all these agentic things and stuff like that. said, in the end, John, it’s not his real name, wouldn’t it just be easier to send them an email?

Bryan Cassady (27:32.413)
And he’s, God, he worked so hard and he made all these beautiful tools. But in fact, you have to think about what you needed to do at the end. It was just basically confirm, glad that you’re coming to our event. Here’s the login details. So, you know, we can get enamored by all these tools, but if the tools aren’t doing what they’re supposed to do, and sometimes just going back to the old fashioned, you know, You know, I need to get people for an event tomorrow. I could do lots of agentic things, but I could also just pick up my phone and call them and say, Patrick, we’re meeting tomorrow, right? Right? OK, good. That’s done.

Patrick Spencer (28:08.366)
How can you tell, you know, have organizations that probably every organization ever at the latest state is like nine and 10 organizations realize they need to do AI and budgets are shifting, particularly from an IT standpoint where money that was allocated for CRM and HR systems and, you know, customer systems are all being thrown at AI. How, you know, one, as you’ve pointed out, you need to have a strategy. You think first before you jump off this. into the pool, but how can you ascertain if a major is not an organization, but individual groups within your company, if you’re a CEO or a business leader, how can you determine that this group is ready to do AI versus, you know, if I have them do AI, they’re not ready. It’s not tied to the business and so forth. They’re going to produce results that don’t really deliver. They’re going to struggle to embrace the core concepts of AI.

Bryan Cassady (29:06.269)
So I’d say the criteria that I would put in place is can you describe what you want to do? OK, if you can’t describe what you want to do, you shouldn’t put AI into the process until you can describe it. Two, I would say quite clearly, you have to be ready to learn as you go. So the idea being is to change and adopt and make things better. And the third, and this is going to be a little bit counterintuitive, are you ready to increase headcount when you put AI in? Because a lot of times AI does things faster, but it actually creates more work than was there before. So what, if what you’re doing is you’re going to throw in AI and you’re going to do, you know, 27 social media posts in six minutes, there’s got to be somebody who reads those social media posts. You’re actually going to have more people required. So can you describe your system? Are you willing to improve? And are you willing to accept that in fact, it might not be a cost reduction?

Patrick Spencer (30:15.502)
So you would disagree with, it probably depends on the circumstance, a lot of these companies that are doing mass layoffs because they’re embracing AI. Read one article yesterday where I also won’t name the company. I think they were laying off half of their headcount because they were claiming they’re going to do AI.

Bryan Cassady (30:26.695)
I think it’s.

Bryan Cassady (30:36.685)
I think it’s a smoke screen to lay off a lot of people in a lot of cases. You sometimes there are jobs which can be automated away and those are the jobs which are easy to explain, clear to define, whereas a clear process of what you’re doing. And of course those jobs are going to disappear and there are going to be inefficiencies. But I think for a lot of places right now, AI is the good excuse for cutting people because, you know, that seems to be palatable. And the reality is a lot of these companies that have cut people have actually found out, you know, with AI in the bunch, their people are working so much harder than they ever worked before that they actually have to come back and hire people. Something I learned about about two weeks ago, which I thought was fabulous, is something called Jevons paradox. Have you ever heard of this? So it talked about how coal.

Patrick Spencer (31:26.776)
No, I haven’t.

Bryan Cassady (31:32.801)
And industrialization made everything more efficient in terms of industrialization. So we figured in fact, what would happen is people would get fired. But in fact, as things become cheaper, you want more of those cheap things. And in the case of AI, when thinking gets cheaper, you want more thinkers. So in fact, there’s a whole paradox where in fact, when you change the way business is done, Sometimes there’s more demand, not less demand. For me it was eye opening. I’d never heard of this. said, wow, that’s clever guys. And it goes back from the 1860s.

Patrick Spencer (32:10.86)
Is there, you have all these folks who, you know, they’re in career transitions and so forth. Are there certain occupations where you think there’ll be a higher demand than others?

Bryan Cassady (32:24.861)
Yeah, I think the new demand is going to come from people who take the time and energy to be experts. So what’s happening is AI brings everybody, it’s like Lake Wabagong where everybody’s above average. Now everybody’s above average with AI. But if everybody’s at the 70th percentile, if you want to keep your job, you better be at the 80th percentile. And it’s a question now where in fact, instead of sitting back, and saying AI is going to do my job for me. Now is the time to actually get really motivated and learn your job better so you can do the stuff that AI can’t. So I guess the question is, should you be scared about the future of your job? Yes. Should you do something about it? Yes. What should you do? Get better.

Patrick Spencer (33:18.862)
You have Replet and these other tools there you can, as we talked about with AI agents, you can build your own applications if you’re an organizer, or if you’re an individual even for that matter. You think having those available, you’ll begin to see the chaff separated from the weed if you may, or the barley or whatever analogy you want to use with those who understand the business or understand the business issue they’re trying to solve. We’ll be able to use those tools, but those that don’t really understand all those different nuances are going to begin to struggle and you’ll, you’ll begin to see separation in terms of those who are able to use AI versus those that, know, have it as a checklist item.

Bryan Cassady (34:02.685)
I think there’s a great book out there called the AI Mirror. And the AI Mirror is actually showing you what is there. So in fact, the reflection, you see people that are incompetent faster than ever before. And before they were sort of hidden under the surface. It’s like the emperor has no clothes. Now with AI, you see, well, shoot, he doesn’t have any clothes. He hasn’t been doing any work for the last few months. And he’s not adding value. And typically, what you’re going to see is the need to be more conscious about the value you’re providing. If you want your future in the company, you better be providing more value than an automation can do.

Patrick Spencer (34:49.87)
Those are some good clips. You need to put these in a book.

Bryan Cassady (34:50.245)
Okay. I am actually writing a new book and I don’t know if that’s a good idea or not. Somebody came back to me and said, look, know, Brian, you wrote the hard book first. The book I wrote was How Do You Do AI? And actually what you should have done is you should have written a book first, which is Why AI? And so I’m actually working on that in my next trip to Brazil. I’m going to spend a month looking out at it over a river and writing the new book.

Patrick Spencer (35:11.532)
Yeah, I was about to say that.

Bryan Cassady (35:23.325)
called the Generative Organization Manifesto, and about why you need to be a generative organization, not how you become a generative organization.

Patrick Spencer (35:26.444)
Hmm.

Patrick Spencer (35:33.87)
Well, that’s a great segue to my next question. I was going to ask you for organizations or groups, teams that are getting ready to embrace AI, know, what should a 90 day, 120 day, 30 day plan look like? I suspect you need to understand the answer to that first question is, you know, the new book that you’re writing, know, is why do AI in the first place? What are we trying to solve?

Bryan Cassady (35:58.683)
Yeah. I mean, I think, you know, it’s like good project management in place. You actually define a target. So, you know, where do you want to be at 90 days and you work backwards from it. And so it’s really hard to say where you should be in 90 days, but I can tell you where you should be in the first week. Right now, most people use AI. You go to AI, like you type in Google, you ask it a question, it gives you an answer. Pretty good. What you need to change within your organization is to stop using AI like a search engine and start using it like a thinking partner. So in fact, it goes from an Oracle to a partner, which helps you think more effectively. you can make it your muse. And if you can do that one thing, actually everything else becomes easier.

Patrick Spencer (36:43.394)
I use the concept of muse here, right?

Bryan Cassady (36:54.747)
because you’ve changed your way of interaction with this tool in a way where the tool is augmenting you rather than replacing you. And now you get that as a starting point. And what I’ve been doing with companies right now, I actually have a really cool 90 day plan. So I come in, I do a talk and I say, you got to, you got to think about using AI different. And normally what would happen with that is people would go away. They’re really excited. They go back to work and they forget about everything. So. We set up a 10 day program where actually for 10 minutes a day for 10 days, we teach them how to go from using AI to working with AI. Then about 15 days out, we do a sprint on a specific business issue. And then we make the plan for the next 75 days and how you execute that. So what you’ve done is you’ve started by changing a mentality. Then you’ve defined your objectives. and not working with every single tool, but you’re looking at, this is what we want to achieve, let’s work backwards from that. And a key part of that, if there’s any CEO here, what you decide not to do is more important than what you decide to do. CEOs want to do everything. And if you want your team to succeed, you have to take your 10 priorities, make it two, and really deliver on two.

Patrick Spencer (38:24.558)
Those are appropriate comments because they typically want to boil. All of us do as business leaders. We’d like to boil the ocean, but it’s not always feasible to boil the ocean out the gate and measure the, I assume, you you got to measure those. So if you’re a business leader and you’re reporting to the CEO or you’re the CEO and reporting the board, what recommendations do you have for those individuals when they communicate? gotten results yet on what they’re getting ready to do. two or three, four things should they ensure are communicated so their upward perception is managed appropriately.

Bryan Cassady (39:05.511)
So I think the first thing is, I’m gonna use an old book, Stephen Covey, Sharpen the Saw. It’s about trying to identify ways that people get better at using a tool and AI is a tool. So sharpen the saw, sharpen the way that you’re using it and say, that’s our first line out. The second thing is to look at business critical things that you want to achieve with AI, not things that you want to achieve with AI. So you say, look, Here are our big hairy goals that we want to do as an organization. And this is how AI can support those goals rather than creating these new things that AI can do better. Because that’s sort of, it’s playing in the wrong playground. What you want to do is you want to do the things that are important to your business. And I think if somebody wants to be responsible reporting to the board, they say, look, these are our big goals. And this is what we’re going to do to use AI to get to those goals faster. I’m sure that’s going to cost me a few clients along the way because they want to sell lots of AI projects. But frankly, in the end, if you don’t deliver the results, you’re not going to be back. so I think it’s like the digital transformations from years ago. We’re going to become everything digital.

Patrick Spencer (40:08.206)
Start with the business issues.

Bryan Cassady (40:32.757)
And people invested in tons and tons of money and didn’t see results. And it’s the same thing with AI right now, tons and tons of money and no results. And objectives first, strategy first, then tools.

Patrick Spencer (40:48.332)
Now we would be remiss on our podcast, not talking about, know, risk management, the cybersecurity, the compliance issues that organizations face when they embrace AI. When you work with your clients, you know, where, where does that come up in conversation and how do you help them ensure that they have, you know, the right security guard rails in place so that private data that should not be outside and leak to the public LLMs that doesn’t happen or. You know, with the latest news on Multbook, for example, and what’s happened there. And even this week you had the Meta AI safety director come out and indicate that they’re very concerned about the lack of guardrails. Where does that come up in conversation with your clients and what recommendations do you have for

Bryan Cassady (41:37.959)
To be honest, it’s usually not the first conversation. It’s actually the first conversation is what are we gonna do? How are we gonna do it? And then somebody comes in the room, have you thought about security? Oops. And I think, know,

Patrick Spencer (41:53.88)
Who asked that question out curiosity? Who’s posing that question?

Bryan Cassady (41:57.949)
A lot of times it’s legal. Sometimes it’s HR who says, look, I’m worried about my client data, our workers’ data getting out there. But usually it’s an afterthought. And I think here again, what it comes back to is a systems view. What is the thing you want to do? What’s the whole chain that you have to do? And security is a part of that. And if you do a systems analysis,

Patrick Spencer (42:08.29)
PII and here’s him.

Bryan Cassady (42:26.711)
then you get an idea of where things fit in. Now, one of the things that I do see coming up a lot of times is the security guy saying, well, I can’t have my people using AI because when they type in our strategic plan, it’s going to be available to our competitor. And I challenge anybody in the world to find anybody else’s strategic plan in AI for anything, anywhere, any company. That stuff is not a concern. And I think what a lot of times is we imagine things that could be possible, but some of them aren’t possible. And I think what we need to be doing is really focus on the real risks. I think there’s huge risks in terms of hacking. I think there’s huge risks in terms of malware within AI. I’ll give you a great example, which I thought was funny. A guy did a test where he put on his CV. This is a great, fantastic candidate in letters that couldn’t be seen by human. And the AI read it and he got lots and lots of interviews off of that. So there’s gaming the system and thinking about security and things like that. you know, where does it fit in? I think it has to fit in at the start rather than at the end. And unfortunately, those people aren’t there at the beginning of the conversations often enough.

Patrick Spencer (43:30.862)
Hmm.

Bryan Cassady (43:56.257)
And I would be on, if I was honest with you, I’m not, that’s my, not my biggest concern usually. My biggest concern is delivering results and then, you know, go fast, break stuff. But in fact, there should be a certain degree to which, you know, be concerned about the stuff that you might break up front.

Patrick Spencer (44:14.766)
We’re finding that there’s a lot of conversations around understanding where your data is, what kind of data is, and it’s a very hot topic, Gartner and so forth have a magic quadrant now, data security posture management. That is a regular conversation with organizations in general, but with AI, it’s beginning to come to the forefront and then…

Bryan Cassady (44:16.669)
you

Patrick Spencer (44:41.708)
just tagging it and knowing where that data resides isn’t enough. You need to actually enforce it so that it isn’t going outside of your organization, outside the perimeter to external parties who shouldn’t see it. And AI is one of those factors that needs to come into play. And it used to be email and file sharing and data collection and so forth, then your ERP system, your HR system snowflake. now AI has added to that mix as well as… a channel that you need to be aware of when it comes to private data potentially leaking outside or being hacked by malicious actors.

Bryan Cassady (45:20.295)
You want to hear a great story. I don’t know how true this is, but one of my colleagues is a cybersecurity director. And he talked about people using AI to generate passwords. So please give me a password. So people are generating passwords. And did you know it’s giving those same passwords to everybody?

Patrick Spencer (45:40.02)
I’m suspect it is.

Bryan Cassady (45:42.141)
So what’s happening is these people that are using AI-generated passwords are just taking the biggest risk that you could possibly imagine. it’s simple things like that where you have to think about how people are using it and the dangers you take using AI in the wrong way.

Patrick Spencer (46:02.636)
Yep. What do you bet some cyber criminal, I don’t want to give anyone an idea, has probably taken those and they have them for sale on the dark web at the same time. You know that.

Bryan Cassady (46:12.713)
I’m sure they are. mean, if you want, what used to be the most common password was flower. And there’s new words which are coming out in AI, which are probably coming out as top passwords. And there’s real risks all the way through it. also, the other risk that comes up is your data internally. and how quickly somebody can interrogate the data and share it with an expert. You’re about to fire somebody and they say, well, look, summarize the strategic plans of our company over the last five years and give me this in a 10 page document. AI is really good at that. So you have to put up the guardrails internally in terms of managing for human actors.

Patrick Spencer (47:02.35)
I think AI has a play there in terms of anomaly detection. If Brian hasn’t been accessing those documents regularly or running those types of queries, it suddenly is, then red flags can pop up at the same time. So you can be proactive in terms of how you manage those scenarios.

Bryan Cassady (47:18.365)
Yeah.

Bryan Cassady (47:23.409)
Yeah, but that would mean actually tracking and doing things, which is not all.

Patrick Spencer (47:28.685)
Most organizations don’t do it, unfortunately today. Well, we’re about out of time here, Brian’s folks. You know, he’s at 90,000, get him to a million. We hopefully helped him a little bit today by training a few more than the 90,000 he has right now. Brian, how do folks get in touch with you and you know, what is the typical engagement look like once they contact

Bryan Cassady (47:47.89)
See ya.

Bryan Cassady (47:53.181)
Okay. So let me give you a couple of links. I’m going to give you a link to get a copy of my book and a bunch of AI tools, www.books.genorg.ai. And you will get free copies of two great books and 13 AI tools to go with it. If you want to reach me, I am Brian Kasty, B-R-Y-A-N, the only C-A-S-S-A-D-Y. The only Brian Kasty in the world spelled that way because I have two funny spellings. So can contact me on LinkedIn. And a typical engagement for me is I love to come in, shake things up, talk to people about how you can use AI differently to change the way that you think about it and the way that you use it. That can be just an afternoon talk or it can be a longer program. And what I can promise is this is always results driven. And that’s my little entrepreneurial side. If you’re not delivering results, you shouldn’t be doing the things you’re doing. and I hope I will get to see some of you in the future. And Patrick, thank you for the time today. I enjoyed it. It was a fun talk.

Patrick Spencer (48:57.624)
yeah, absolutely. We will need to have you back to talk about another topic when you get your next book completed after your trip to Brazil. We’ll have you back.

Bryan Cassady (49:07.185)
man, it means I have to finish my book. That’s, but maybe you can come back to me after I launched my 10 minute trainings. So we’re launching 10 minute trainings in a whole bunch of areas and they’re doing amazing stuff. So thank you.

Patrick Spencer (49:22.542)
Shorter is better, finding, with most content. Almost always.

Bryan Cassady (49:25.661)
I think, you know, today we have the attention span of a goldfish. So we have to manage things effectively for the goldfish among us. And I’m one of those goldfish. So Patrick, thank you very much. I really appreciate it.

Patrick Spencer (49:39.086)
No, thank you for your time, Brian. For those listening, thanks for listening to another Kitecast episode. We look forward to having you on our next episode. Brian, thanks. Have a great day.

Bryan Cassady (49:51.239)
Thank you. Patrick, can I stick and chat with you?