Latest / Matt Brown Show - Conversations That Power The Business World. / MBS993 - The AI Advantage: Why You Still Need a Human at the Helm with Drew Marshall

Transcript

MBS993 - The AI Advantage: Why You Still Need a Human at the Helm with Drew Marshall · Matt Brown Show - Conversations That Power The Business World.

• 0:05
  SPEAKER_00: Hi everybody and welcome back to yet another cracking edition of the Matt Brown show.
• 0:09
  SPEAKER_00: This is the AI Advantage Series where we are talking to leaders all about how AI is going to shape the future of their industry.
• 0:18
  SPEAKER_00: And with us in the hot seat today is none other than Drew Marshall.
• 0:21
  SPEAKER_00: He is the chief operating officer of Kipna Trigo, I believe.
• 0:27
  SPEAKER_01: Well done, Matt.
• 0:28
  SPEAKER_01: Well done.
• 0:30
  SPEAKER_00: So you guys are doing some awesome stuff in critical thinking.
• 0:32
  SPEAKER_00: You've, you know, as a firm, you've been around for uh almost over five decades, I believe.
• 0:37
  SPEAKER_00: Uh, but what is the elevator pitch for our audience around the world today, Drew?
• 0:41
  SPEAKER_01: Yeah, so I think one of the things that that everybody recognizes is that problem solving and decision making are those skills that everybody needs to be successful in every aspect of their life.
• 0:51
  SPEAKER_01: And we've been in this space for for quite a long while.
• 0:54
  SPEAKER_01: And one of the things that we recognize is that the ability to transfer good critical thinking skills into the people in your organization is one of the fastest ways you can get a return on your investment and drive performance.
• 1:06
  SPEAKER_01: So for us, it's it's really interesting at this stage because our focus on helping people perform at their best more often than not is kind of one of the reasons why we exist as a business.
• 1:19
  SPEAKER_01: Uh, you know, our founders came together to focus on best practice critical thinking skills and develop systems to translate that into effective practice in organizations.
• 1:28
  SPEAKER_01: And that's something that carries us to today.
• 1:30
  SPEAKER_01: And uh I think the interesting thing for us is when we look at the technology that we are uh surrounded by, living with, swimming in in some cases, uh the need for critical thinking skills effectiveness has never been greater.
• 1:46
  SPEAKER_00: So, how do you define critical thinking today?
• 1:48
  SPEAKER_00: What does that actually mean?
• 1:50
  SPEAKER_01: Yeah, so critical thinking is the ability to gather, sort, and organize information in a way that enables you to more effectively resolve the concern that you have.
• 2:00
  SPEAKER_01: When we think about critical thinking skills, what we're focusing on are those things that cause an organization to pause or in some cases get stuck.
• 2:09
  SPEAKER_01: So uh if we're not clear about the direction we should take, then prioritization requires critical thinking skills.
• 2:16
  SPEAKER_01: Uh, if you have a problem that needs to be solved before you can advance the business, then to solve for root cause requires a set of critical thinking skills.
• 2:25
  SPEAKER_01: And uh charting a f a path forward as a leader requires you to think about the strategy of the organization.
• 2:30
  SPEAKER_01: And that's another set of critical thinking skills in and around decision making.
• 2:35
  SPEAKER_01: So all of those different frames, prioritization, problem solving, decision making, risk mitigation, require you to have a really clear understanding of the concern you're trying to address, not get it confused with another kind of concern, and make sure that as you apply those critical thinking skills, you're advancing the most urgent or impactful needs of the business.
• 2:56
  SPEAKER_00: Yeah, it's interesting, right?
• 2:57
  SPEAKER_00: And so I think this whole idea of critical thinking is going to become even more uh acute and more important, right, as we kind of move into this AI world.
• 3:07
  SPEAKER_00: Uh, you were sharing a story with me as an example because you guys have such an incredible body of work across multiple um, you know, verticals that um hallucinations do occur, right?
• 3:19
  SPEAKER_00: So you can have the best data in the world, uh, but if you if a human is not able to apply critical thinking to the output that the AI is generating, uh then that's gonna cause uh concern for you know uh productivity outcomes, all that kind of stuff.
• 3:36
  SPEAKER_01: Yeah, yeah.
• 3:37
  SPEAKER_01: One of the phrases that I heard probably two years ago now was from a person who's uh doing a lot of uh deep work in this domain uh and uh is um a university professor and researcher um up in New York.
• 3:54
  SPEAKER_01: And um Connor Grennan's his name.
• 3:56
  SPEAKER_01: And Connor Grennan was talking about the fact that when you consider AI in the organization, uh you can't think of it as something that's standalone.
• 4:05
  SPEAKER_01: Even if you're automating processes or using AI to augment, you always have to have what he referred to as a human at the hell.
• 4:12
  SPEAKER_01: And that is, I think, an area of criticality for most organizations as they explore integrating AI into their operations, is you have to have people who understand your business explicitly if you're going to start automating aspects of your business.
• 4:27
  SPEAKER_01: Because if you don't, what will happen is mistakes will be made.
• 4:31
  SPEAKER_01: You will create tomorrow's problems today by doing things with AI that you don't understand the limits of, or potentially um compounding already existing process or systems problems that you have.
• 4:44
  SPEAKER_01: So when we look at our body of knowledge, which is around the construction of the approach to gathering, sorting, and organizing information in order to analyze it effectively, the thing about AI or AI in terms of large language models is that the data that you use to train that AI creates the natural limits for its performance.
• 5:06
  SPEAKER_01: But if you don't reinforce those limits, you end up in a place where the AI itself will extrapolate from the data that's available to it and kind of skate off into hallucination.
• 5:17
  SPEAKER_01: One of our clients, uh large tech company, was applying uh AI and the use of our processes uh to their body of knowledge.
• 5:26
  SPEAKER_01: So they're a company that has a large number of software products and services that they bring to market.
• 5:31
  SPEAKER_01: And they were trying to use Ketna Trigos problem analysis methodology in the context of AI and their body of knowledge.
• 5:39
  SPEAKER_01: And what they discovered was that in the Venn diagram, they had uh KT as a process, they had their body of knowledge, but they didn't have a human who was basically looking at the way they were being integrated and was vetting the common sense of the outcomes that were coming out of AI.
• 5:58
  SPEAKER_01: And what ended up happening was they had a very, very high, uh I think it was high 60 percentile um uh uh arriving at hallucinations as a result of the work that they were doing, and they had to abandon that initial uh approach.
• 6:12
  SPEAKER_01: Subsequently, they'd been able to modify their approach, they've got that uh that hallucination hallucination down to single digits.
• 6:21
  SPEAKER_01: But the caution there was we were talking about an organization who has a whole bunch of data scientists at its uh beck and call.
• 6:28
  SPEAKER_01: It has a very deep bench strength in terms of its understanding and application of technology.
• 6:34
  SPEAKER_01: For the run-of-the-mill organization out there, your manufacturing concern, your your uh your local bank, if they're if they're starting to apply AI, my question is going to be so do you understand the limits of what you're trying to do?
• 6:48
  SPEAKER_01: And how are you controlling for those limits by the human interaction that you could that you need to have in place in order to be successful?
• 6:55
  SPEAKER_00: So how do you know where those limits are?
• 6:57
  SPEAKER_00: I mean, if they're constrained by the the data set that the LLMs are being trained on, where does where do the constraints start to manifest?
• 7:08
  SPEAKER_01: Yeah, so I think one of the things that people have started to recognize is that as they're training, they have to train on data that is already known and outcomes that are known in order to be able to validate whether or not AI is actually arriving at a common sense response.
• 7:25
  SPEAKER_01: I think where we're seeing um really great benefit for AI and its application in business is where it's helping people push the envelope in terms of uh uh creativity or push the envelope in terms of extending their range of understanding of what might be.
• 7:41
  SPEAKER_01: So when you think about decisions and you think about the approach to decision making, AI is really beneficial in helping you figure out the range of objectives that you might need to consider when you're making a decision or a recommendation.
• 7:52
  SPEAKER_01: It's also very useful in helping you realize or extend your understanding of the range of alternatives that you might select from.
• 8:00
  SPEAKER_01: At the end of the decision-making process in terms of risk mitigation, it's also good for helping populate the set of risks that you might need to consider.
• 8:08
  SPEAKER_01: But if you're talking about a situation where there is a process which has broken, or something mechanical that has broken, and there is a set of understanding in and around that that is bounded by the system itself and the environment that it's operating in, one of the challenges is when left to its own devices, AI will struggle to identify true root cause.
• 8:33
  SPEAKER_01: So a number of things need to come into play.
• 8:36
  SPEAKER_01: What's the human performance system that exists over and around your AI system from an intervention perspective that provides oversight and provides guidance and input?
• 8:45
  SPEAKER_01: So there's early detection in terms of uh the early stages of the hallucination.
• 8:50
  SPEAKER_01: The other is what's the way in which you can frame AI so that it actually is self-checking?
• 8:56
  SPEAKER_01: And that's where we're seeing a lot of progress being made on AI, which is clearly defining the knowledge boundaries and saying if you get to a point where there is no knowledge in your large language model, then you do not create anything else and you offer up this is the extent, this is the boundaries, boundary state.
• 9:14
  SPEAKER_01: And I think that's where we're going to start seeing not necessarily self-awareness on the part of AI.
• 9:20
  SPEAKER_01: I think what it's going to end up at is the end of its processing limits.
• 9:23
  SPEAKER_01: It's coming to the end of a knowledge set and saying, yep, I don't have any more information at here.
• 9:28
  SPEAKER_01: I need more data.
• 9:30
  SPEAKER_01: And I think that message alone in the application of AI is probably an area of criticality because then that's where the human picks up and starts to supplement that that boundary set with a new set of data or begins the process of research where they've reached the understanding of the AI and the human understanding.
• 9:47
  SPEAKER_01: We're now starting to go into pure roos, uh pure research domains.
• 9:52
  SPEAKER_00: So how have you, as obviously being a business that applies critical thinking or works with companies to apply that sort of stuff, how have you applied AI in uh KT today?
• 10:04
  SPEAKER_01: So one of the things that's really important for us is making our thinking visible.
• 10:09
  SPEAKER_01: So to get to the point where AI is becoming a fundamental part of the way we do business, what we're doing is questioning the whole business from top to bottom.
• 10:20
  SPEAKER_01: So one of the first questions that we have as a leadership team is can AI be applied in this domain?
• 10:26
  SPEAKER_01: Can we automate using AI?
• 10:27
  SPEAKER_01: Can we augment using AI?
• 10:29
  SPEAKER_01: So the key for us is defining where and how to experiment.
• 10:34
  SPEAKER_01: The other thing for us is prioritization.
• 10:37
  SPEAKER_01: Um, you can try and upend your entire business, your entire business model through the practical application of AI.
• 10:44
  SPEAKER_01: But if you don't have an understanding of the implications of the testing that you're doing, you can break things and not necessarily have good ways to recover.
• 10:53
  SPEAKER_01: So we're very um we give a lot of forethought to the way in which we're running those experiments.
• 10:59
  SPEAKER_01: We document them, we manage them, and we're reporting them out to the organization overall.
• 11:04
  SPEAKER_01: Um, the work that we do uh with our clients uh is in many senses something that is very private between us and the client.
• 11:13
  SPEAKER_01: We operate under very deep confidentiality.
• 11:16
  SPEAKER_01: The other thing is we operate under almost a Hippocratic oath with our clients, which is first, do no harm.
• 11:22
  SPEAKER_01: So as we're exploring this technology with ourselves, uh, we're very conscious of the limits of what we can and should do with our clients.
• 11:30
  SPEAKER_01: In fact, many clients have stipulations in their contracts with us about the extent to which AI should be present in the work product that we're providing.
• 11:38
  SPEAKER_01: So that's an ongoing negotiation, and we need to, as the service provider, be really clear about what those boundaries are.
• 11:44
  SPEAKER_01: Um, for organizations that build physical products, it's going to be really interesting to see what the implications are for them in terms of um rapid onboarding of new vendor partners.
• 11:55
  SPEAKER_01: We're looking at a situation in uh in North America in particular, with the economic conditions being as they are.
• 12:02
  SPEAKER_01: A lot of uh reshoring and onshoring is happening to supply chains.
• 12:07
  SPEAKER_01: I'm seeing AI being applied in those situations where the rapid identification of vendor partners is taking place, the evaluation of those vendor partners, and then the onboarding of those vendor partners is really collapsing as a result of AI being applied appropriately in those conditions.
• 12:23
  SPEAKER_01: So there are lots of significant opportunities for the judicious use of AI.
• 12:29
  SPEAKER_01: The other thing that I think many organizations need to be conscious of is AI is not a panacea.
• 12:36
  SPEAKER_01: It's not a it's not a pill you can take and it's going to solve all your problems.
• 12:40
  SPEAKER_01: The way to think about it is it's um the judicious use of a tool.
• 12:45
  SPEAKER_01: So you wouldn't use a hammer to unscrew a Phillips head screw.
• 12:49
  SPEAKER_01: So think of AI as serving a particular purpose for a particular need and serving for that need.
• 12:57
  SPEAKER_01: One of the things we're we're seeing is a lot of clients are trying to single source their AI solutions.
• 13:04
  SPEAKER_01: And we would suggest, from a decision-making perspective, that's uh potentially increasing the amount of risk that you're choosing to bear as an organization.
• 13:13
  SPEAKER_01: And we would probably caution you to consider going with a much broader approach to the application of AI across the enterprise.
• 13:19
  SPEAKER_01: It doesn't mean that you should be trying everything all at once.
• 13:22
  SPEAKER_01: Uh, what it means is you should be really judicious about where and how you're choosing to use AI to the benefit of customer value generation.
• 13:31
  SPEAKER_00: Absolutely.
• 13:32
  SPEAKER_00: So, Drew, one of the things I found quite interesting from what you just described was that your clients are specifying to what extent uh AI can be present in the solution set, right?
• 13:43
  SPEAKER_00: Um, and so what that what I hear when I hear something like that is that there's a skepticism about AI or there's a uncertainty about how to adopt it and how fast to adopt it and those sorts of things.
• 13:56
  SPEAKER_00: Uh is that is that a fair characterization?
• 14:00
  SPEAKER_01: Yeah, I I think as with all new technology, Matt, I think one of the things that's that's interesting is this is this is a a playbook that we've seen previously.
• 14:09
  SPEAKER_01: There, you know, we talk about the adoption curve.
• 14:11
  SPEAKER_01: There's always the the you know cutting edge or bleeding edge of uh you know adopters, then there's the the general mass of adoption, and then there's the long tail of adoption at the back end of it.
• 14:21
  SPEAKER_01: And you know, name a uh technology that's been introduced into the enterprise in you know the last you know two, three generations, they all have a very similar pattern.
• 14:30
  SPEAKER_01: I think what also happens when we see these new technologies being brought to bear in in business is that the time frame between earliest adoption and lagging adoption is starting to collapse.
• 14:42
  SPEAKER_01: And that's the same with the adoption of any technology, the mobile phone, you know, moving on onto iPhones, etc.
• 14:49
  SPEAKER_01: That that collapse of adoption is happening in scale.
• 14:54
  SPEAKER_01: The issue with AI is people are still trying to figure out where it fits within their business model.
• 15:01
  SPEAKER_01: So I think what you're going to see is AI is going to start um causing organizations to try to dismantle existing business models without the benefit of having a good landing place to go to when they actually do implement AI.
• 15:16
  SPEAKER_01: Um, it's already up-ending uh pretty significant parts of business processes in organizations.
• 15:22
  SPEAKER_01: For example, in the legal profession, um, there's a pipeline of new junior associates that come in to work in law firms, and that pipeline is predicated on being able to put those uh new uh new uh baby lawyers, so to speak, onto um research programs or onto discovery cases.
• 15:42
  SPEAKER_01: Well, all of that work can be done now with AI.
• 15:45
  SPEAKER_01: So the number of people required to feed that funnel gets fewer and fewer and fewer over time.
• 15:51
  SPEAKER_01: The problem, though, is a law firm is predicated on a pyramid structure where it's got a really wide base of junior attorneys, and all of them working towards being senior partners in the business.
• 16:02
  SPEAKER_01: Well, if you take out that bottom layer, what does that business model start to look like now?
• 16:08
  SPEAKER_01: On top of that, that process of discovery and research at the junior level is actually the way lawyers engage with the law and become very familiar with the way in which law can be used on behalf of a client and uh, you know, both the plaintiff and counsel.
• 16:25
  SPEAKER_01: So the challenge for law firms is not only do they not have a need for these lawyers, but they now don't have a pipeline of people who can move up in the pyramid.
• 16:33
  SPEAKER_01: So it's it's going to cause some really fundamental questioning of business models that have been in place uh for a long, long period of time.
• 16:42
  SPEAKER_01: Uh consulting firms, I'm in training and development and consulting.
• 16:46
  SPEAKER_01: Um I'm looking at our business and I'm looking at you know larger consulting firms in uh in the space, and they are going to be struggling with that similar prospect.
• 16:55
  SPEAKER_01: You know, the the analysis, the research that happens uh amongst those junior personnel, that's no longer going to be required.
• 17:03
  SPEAKER_01: But also that training process is how they become familiar with serving clients, engaging with clients, developing the human relationship skills that are necessary to be a senior partner in a firm.
• 17:15
  SPEAKER_01: So uh I think just the service sector alone is running smack into the implications of AI.
• 17:21
  SPEAKER_01: But I think every business that you can name is in some fashion going to be wrestling with AI, much like every business had to, in some fashion, wrestle with the implementation of the World Wide Web and the Internet.
• 17:32
  SPEAKER_01: Um, you know, everybody was like, well, what actually is the internet?
• 17:36
  SPEAKER_01: I you know, I can remember when I worked many years ago at a very large software firm, we had one of the first corporate internet uh websites.
• 17:44
  SPEAKER_01: And there was a presentation to us all, and we were all looking at it, I was like, okay, it's interesting, I guess.
• 17:51
  SPEAKER_01: And I you click there and it goes somewhere else.
• 17:54
  SPEAKER_01: Okay.
• 17:54
  SPEAKER_01: Well, you know, now it's in an indispensable part of everybody's life on a daily basis.
• 17:59
  SPEAKER_01: So AI in 20 years will probably be talked about in the same way.
• 18:04
  SPEAKER_01: It's an indispensable part of everybody's daily business life and daily life.
• 18:09
  SPEAKER_00: So you're in the consulting uh space, as you mentioned, training and developments and consulting.
• 18:14
  SPEAKER_00: Um how do you how do you see the role of consultants, you know, changing now that everybody has access to all the information in the World Wide Web through a GPC interface as an example?
• 18:26
  SPEAKER_01: Yeah, I look, I think the role of trusted advisor is going to continue.
• 18:33
  SPEAKER_01: I think the way in which you become a trusted advisor may change and evolve over time.
• 18:39
  SPEAKER_01: Um being at the top of a body of knowledge is what made you a trusted advisor previously.
• 18:46
  SPEAKER_01: You have a deep understanding of tax law, or you have a deep understanding of manufacturing processes, or you have a deep understanding of advertising and marketing.
• 18:54
  SPEAKER_01: Um, I think that understanding, that that approach to developing a body of knowledge, um, it is almost impossible for a single person to sit on top of a body of knowledge and own that entirely.
• 19:08
  SPEAKER_01: But being able to help clients navigate bodies of knowledge, being able to integrate different aspects of bodies of knowledge and remain um relevant and remain viable and remain trusted because the advice you offer is trustworthy, I think that is going to remain a critical need.
• 19:29
  SPEAKER_01: I think senior leaders have already found they don't have enough hours in the day to stay on top of things.
• 19:35
  SPEAKER_01: And I think the people that report to them are in the same boat, and the people that report to them are in the same boat.
• 19:41
  SPEAKER_01: So I think there will always be a need for effective consultants who are working on behalf of their clients, value generation.
• 19:50
  SPEAKER_01: I think what you're going to see go away is um driving consulting offerings to the lowest possible level of skill in an organization in order to retain your billable margin on projects, because people are going to be able to see right through that.
• 20:08
  SPEAKER_01: They're going to be able to see the gap between what you've asked AI to do and your understanding of what you're presenting.
• 20:15
  SPEAKER_01: And this is one of the things that I see happening a lot in academia where students who are going through universities are using AI to produce quite good work because the AI is using a body of knowledge that they're tapping into.
• 20:28
  SPEAKER_01: But when asked critical questions about what it is they're presenting, they fall short.
• 20:34
  SPEAKER_01: And I think that first step where you're being queried on what you're presenting about whether or not you do understand it, is where you're going to see a lot of trust frayed.
• 20:42
  SPEAKER_01: And in fact, it will be broken in some cases.
• 20:45
  SPEAKER_01: And I think it's going to be very difficult for firms to recover in those situations.
• 20:49
  SPEAKER_00: Yeah.
• 20:50
  SPEAKER_00: Um, very quickly, I want to talk to you about ethics.
• 20:53
  SPEAKER_00: Um, this is obviously a very important thing, right?
• 20:56
  SPEAKER_00: I mean, if you think about consulting and AI and then, you know, trust being the bridge, you know, between uh making an impact or not.
• 21:06
  SPEAKER_00: Um, how do you see the role of ethics playing out in the adoption of AI?
• 21:12
  SPEAKER_00: Because there's one thing getting like I keep talking about this idea, like we're all stuck inside the bottle of our own problems, we can't read the label.
• 21:21
  SPEAKER_00: And so that's where consultants have always historically added a lot of value, meaning they can give you perspective so that you can read the label again.
• 21:30
  SPEAKER_00: Um, and so now with AI and the accessibility to AI in terms of knowledge base, sets, all that kind of stuff, um what role does ethics now play in this whole engagement between clients and AI and a consulting uh firm?
• 21:50
  SPEAKER_01: I look, I think uh ethics is of paramount importance because I think you know that lens lends itself to the establishment of a trusted.
• 22:00
  SPEAKER_01: Advisor relationship and credibility.
• 22:03
  SPEAKER_01: When you think of trust as an equation, it is your capability and your credibility over the risk of you engaging in what you will do.
• 22:13
  SPEAKER_01: So, you know, capability, can you do something?
• 22:15
  SPEAKER_01: Credibility, will you do something?
• 22:18
  SPEAKER_01: And risk, did you or did you not do that?
• 22:22
  SPEAKER_01: And when you think about that equation, at every step of the way, if you think about AI taking on that responsibility, I think that's that's essentially trying to delegate to a dumb system.
• 22:34
  SPEAKER_01: And dumb meaning there is not critical thinking going on in the AI.
• 22:38
  SPEAKER_01: It is not thinking about morals.
• 22:40
  SPEAKER_01: It's not thinking about the boundary states of particular emotional conditions or psychological conditions.
• 22:49
  SPEAKER_01: What it is doing is relying on a data set and huge amounts of processing power to review that data set and come back with its assessment of what you need at a particular point in time based on your prompt or your query.
• 23:03
  SPEAKER_01: Ethics is the way in which we navigate the political, emotional, and psychological landscape around us.
• 23:11
  SPEAKER_01: When you think of doctors, for example, doctors and lawyers, professions have particular sets of ethics and ethical conditions that they have to adhere to.
• 23:22
  SPEAKER_01: And I think the same goes for consultants.
• 23:24
  SPEAKER_01: If you are going to be a consultant, you need to practice consulting from an ethical perspective, which is you are providing knowledge based on your first hand experience or the experience of your firm to a client for which they are paying.
• 23:40
  SPEAKER_01: So the boundary set there is, you know, there has to be a human in the mix that is managing that ethical state, managing those ethics.
• 23:49
  SPEAKER_01: I think where we're going to come on, uh come apart in places is where people are so rapidly trying to implement AI, they're not considering the ethical implications of the AI that they're implementing, both for their own organization and for the for the clients that they're trying to serve.
• 24:08
  SPEAKER_01: So I think for for us, the caution for us as consultants is to think about the what-if conditions.
• 24:16
  SPEAKER_01: We have a process called potential problem analysis at Kep Natrigo.
• 24:20
  SPEAKER_01: And um, you know, potential problem analysis asks the fundamental question when we do this, what could go wrong?
• 24:27
  SPEAKER_01: And when we are working with clients, that's one of the operating mindsets for us.
• 24:32
  SPEAKER_01: So as we're offering client, this client this advice or offering this client this uh process improvement or system improvement, what could go wrong?
• 24:40
  SPEAKER_01: What are the things that we need to consider and either prevent from going wrong, prevent the cause of those actions, or have contingencies in place?
• 24:48
  SPEAKER_01: And we're very upfront with our clients that this is the way we work.
• 24:52
  SPEAKER_01: We use our own processes in our application of our own methods with our clients.
• 24:57
  SPEAKER_01: And they should expect that from us.
• 24:59
  SPEAKER_01: That's an that's a part of the way we ethically frame our approach to work.
• 25:03
  SPEAKER_01: We Microsoft used to call it eating your own dog food.
• 25:06
  SPEAKER_01: We do that on a regular basis.
• 25:08
  SPEAKER_01: But I think there is a tendency when you think about new technology to to want to color outside the lines as a way of getting ahead quickly.
• 25:18
  SPEAKER_01: And I think that's that's I think where we're going to see some pretty significant ethical shortcomings.
• 25:24
  SPEAKER_00: Can agree with you more.
• 25:26
  SPEAKER_00: Drew, um, one more fun question for you.
• 25:28
  SPEAKER_00: So forget consulting and everything we've discussed.
• 25:30
  SPEAKER_00: Sure.
• 25:31
  SPEAKER_00: But if you were to predict one fun or surprising way that you think AI is going to affect everyday life in the future, what would you say that is?
• 25:44
  SPEAKER_01: Um I think you know, it's it's interesting when you think about the ways people connect today.
• 25:57
  SPEAKER_01: Like if you are single and you're looking for somebody, there are dating apps or there are networking events.
• 26:03
  SPEAKER_01: I think it will be very interesting to see in that realm how AI is applied so that people are meeting each other's needs more successfully, not necessarily putting on a front or being someone that they're that they at heart are not.
• 26:19
  SPEAKER_01: But I think um it'll be really interesting to see if AI can help people more authentically make connections.
• 26:26
  SPEAKER_01: I think that's that's for me, it's it's we people keep forgetting that you know humans are already in the mix with AI.
• 26:33
  SPEAKER_01: I think you're going to see it being pushed in all sorts of boundary, all to all sorts of uh different boundary conditions.
• 26:39
  SPEAKER_01: But that for me was just uh something something meant, someone mentioned it in passing the other day.
• 26:43
  SPEAKER_01: It's like, you know, when when AI AI is in dating apps, we're all going to basically be looking at you know perfect versions of ourselves.
• 26:50
  SPEAKER_01: I said, I don't think that's the case.
• 26:51
  SPEAKER_01: I think there's a there's an opportunity to actually discover what makes people tick more effectively through practical use of AI, and I think that's gonna make for more powerful connections over time.
• 27:01
  SPEAKER_01: So it'll be interesting to see.
• 27:03
  SPEAKER_00: Yeah, it will be.
• 27:04
  SPEAKER_00: I was watching uh the imitation game with Ben and Batch.
• 27:08
  SPEAKER_00: And uh he there's this um I'm sorry, it was like the accountant two.
• 27:13
  SPEAKER_00: Um I don't know if you've seen it yet, but you've got Ben Affleck there and he goes to this dating uh event and basically re-re-engineered all their algorithms to find better matches and all that kind of stuff is kind of quite a weird thing.
• 27:27
  SPEAKER_00: Having said that, obviously AI will do better than humans at that.
• 27:31
  SPEAKER_01: Well, I think you know everybody complains that humans are so fallible or humans, humans uh you know, they're not as authentic as they should be.
• 27:39
  SPEAKER_01: But the problem is when people feel like they're being authentic, they don't feel like they're seeing because they're not as bright and shiny as the next person.
• 27:47
  SPEAKER_01: So I think there's I think there's a pretty significant opportunity to find those matches, like the needle in the haystack matches that people were looking for.
• 27:54
  SPEAKER_00: Yeah, 100%.
• 27:56
  SPEAKER_00: 100%.
• 27:56
  SPEAKER_00: Drew, thank you so much for being on the show.
• 27:58
  SPEAKER_00: Um, I think this whole critical thinking story is going to become actually more important, and I suppose for you, even as a firm, it's gonna unlock a bigger opportunity uh with this AI wave uh that we're inheriting.
• 28:11
  SPEAKER_00: So, yeah, very exciting for you guys, and uh obviously wishing you and the team there at uh KT all the very best of the future.