رونوشت کامل How to Use AI Without Losing Human Judgment: AI Decision-Making and the Future of work | Rana Gujral

software anymore. 00:04:51.519 --> 00:04:53.759 That's cognitive infrastructure. 00:04:53.840 --> 00:04:59.759 And the thing nobody's pricing correctly is what does that do to your organization over time? 00:05:00.000 --> 00:05:11.600 Because a system that sits inside your judgment either builds better instincts in your people or it produces a very convincing stream of outputs that feels like thinking without actually being thinking. 00:05:11.759 --> 00:05:14.879 I mean, those are two completely different companies five years out. 00:05:15.039 --> 00:05:21.279 One has sharpened its taste and the other has quietly outsourced it and doesn't know yet. 00:05:21.759 --> 00:05:22.399 Interesting. 00:05:22.720 --> 00:05:26.480 So you you call the title of the book uh is The AI Instinct. 00:05:26.639 --> 00:05:37.840 Now, instinct is something that gets the it's a biological that gets developed over uh years, especially in business, of uh an individual doing the thing. 00:05:38.240 --> 00:05:42.639 Why do you think that instinct is the right word for software like AI? 00:05:42.879 --> 00:05:48.399 And what would have been wrong with calling it maybe judgment or intuition instead? 00:05:49.440 --> 00:05:55.360 Instinct is the right word because it's the layer underneath judgment and intuition. 00:05:55.519 --> 00:05:57.519 You know, judgment is deliberate. 00:05:57.679 --> 00:05:59.279 I mean, intuition is felt. 00:05:59.360 --> 00:06:03.039 Um instinct is what fires before either of them shows up. 00:06:03.279 --> 00:06:09.919 It's the precognitive pull that makes you turn your head, flinch, trust someone, or distrust someone. 00:06:10.080 --> 00:06:14.240 And that's exactly the layer AI is now operating at in our lives. 00:06:14.399 --> 00:06:14.639 Right? 00:06:14.799 --> 00:06:18.559 So think about what happens when you open your phone, right? 00:06:18.720 --> 00:06:25.279 Before you've made a judgment about what to read, before you've had an intuition about what matters today. 00:06:25.360 --> 00:06:28.000 Um a system has already shaped what's in front of you. 00:06:28.159 --> 00:06:34.159 I mean, it's acted on you at the instinct layer, it's decided what your attention would snap to. 00:06:34.399 --> 00:06:36.320 That's not judgment territory. 00:06:36.399 --> 00:06:38.399 That's older than judgment, right? 00:06:38.559 --> 00:06:40.000 And there's a second reason. 00:06:40.159 --> 00:06:40.399 Yeah. 00:06:40.559 --> 00:06:46.559 Instinct in humans is what you get when experience compounds into something automatic, right? 00:06:46.639 --> 00:06:51.840 So a trader who's been through three crashes doesn't um reason her way out of the fourth one. 00:06:51.919 --> 00:06:54.559 I mean, her body knows before her brain catches up. 00:06:54.720 --> 00:06:59.440 That's decades of um, you know, consequence baked into a reflex. 00:06:59.679 --> 00:07:13.759 The interesting question with AI is whether these systems are developing something analogous, like a compressed residue of um everything they've been trained on that fires as kind of a machine reflex, um, not really reasoning, right? 00:07:13.919 --> 00:07:21.440 Not intuition in the felt sense, something more primitive and I and I think a lot more powerful. 00:07:21.679 --> 00:07:28.960 So if I'd called it the AI judgment, I would have sounded like a book about decision-making frameworks, right? 00:07:29.039 --> 00:07:39.759 So judgment implies uh deliberating self, like uh weighing options, and that's not what's actually happening inside these models or you know, inside the humans using them, frankly. 00:07:39.919 --> 00:07:48.240 So intuition would have been closer, but intuition still lives in the domain of feeling, and it's specifically human, at least, you know, so far. 00:07:48.399 --> 00:07:51.440 And it carries a warmth I didn't want to project into software. 00:07:51.600 --> 00:07:55.440 So I think instinct is colder, it's mechanical in a way. 00:07:55.519 --> 00:08:06.879 I mean, it's what happens when a system, uh, biological or otherwise has absorbed enough pardon that response uh, you know, proceeds through, uh, precedes thought, right? 00:08:07.039 --> 00:08:12.000 And so that's the layer where AI is quietly, I guess, reorganizing human life. 00:08:12.080 --> 00:08:17.519 And it's the layer where if you're not careful, we stop noticing that our reflexes aren't fully ours anymore. 00:08:17.679 --> 00:08:19.600 So I think that's why the word matters. 00:08:19.679 --> 00:08:22.720 It names the place the action is actually happening. 00:08:23.039 --> 00:08:24.399 Yeah, I love it. 00:08:24.720 --> 00:08:30.879 So a formula that you reference in the book is human plus tools plus rules. 00:08:31.120 --> 00:08:39.120 Now, most of the companies that we work with, they've definitely got the first two, but they don't necessarily, I don't even know if they need to have the third. 00:08:39.440 --> 00:08:58.559 In in your lexicon, your context, what does a rule look like that uh, you know, a company that was had 50 people, let's say, using AI today, what would something, what would your definition of that rule, that third part of this mechanism, look like for just a traditional organization that maybe wasn't necessarily software? 00:08:58.879 --> 00:09:02.000 So role clarity, I mean, I think that's where it starts. 00:09:02.159 --> 00:09:05.600 And um, to be honest, I mean, most deployments completely skip it, right? 00:09:05.679 --> 00:09:20.000 So when a human and AI system work on a decision together, uh you have to be able to answer out loud before anything ships who sets the aim, who supplies the evidence, who has the veto, and who gets blamed when it goes wrong. 00:09:20.159 --> 00:09:25.519 So if you can't answer those four questions in a sentence each, you don't really have a deployment. 00:09:25.679 --> 00:09:28.159 You have an accident waiting for a press release. 00:09:28.319 --> 00:09:31.519 Um I'll give you in a failure mode I see

most often, right? 00:09:31.600 --> 00:09:33.919 It's called a moral crumple zone. 00:09:34.000 --> 00:09:47.839 I mean, something goes wrong in a human AI system and the blame collapses into the onto the human operator, like the nurse, the loan officer, the driver, the analyst, who actually had very little control over what the model did. 00:09:48.000 --> 00:09:49.519 The AI is a black box, right? 00:09:49.600 --> 00:09:57.439 I mean, the org chart says the human is accountable, and so the human absorbs the fault for a system they couldn't meaningfully steer. 00:09:57.679 --> 00:10:00.959 You know, responsible deployment means you design that out. 00:10:01.120 --> 00:10:06.399 You don't put a human in a chair, hand them uh model output, and call their signature oversight. 00:10:06.799 --> 00:10:07.759 That's just theater. 00:10:07.919 --> 00:10:12.559 The second thing, and I think this is quieter, is uh automation bias, right? 00:10:12.639 --> 00:10:17.519 When people work next to a confident system long enough, they stop deferring. 00:10:17.600 --> 00:10:19.839 I mean, they sign off, they stop checking. 00:10:20.000 --> 00:10:26.000 And then the signal they used to have, the thing that let them catch the model's uh mistakes, uh, are trophies. 00:10:26.079 --> 00:10:27.439 It just are trophies, right? 00:10:27.519 --> 00:10:32.559 So Aviation figured this out decades ago and called it automation complacency. 00:10:32.720 --> 00:10:37.679 You know, uh pilots who fly an autopilot for years struggle when they have to hand fly through a crisis. 00:10:37.839 --> 00:10:43.759 I mean, the same thing is happening right now in medicine, in credit, in hiring, in content moderation, and we're not naming it. 00:10:43.919 --> 00:10:57.439 So in practice, audit trails you can actually read, um, supervised autonomy, meaning the model runs, but uh human can override, um, and uh, you know, and the override is cheap and fast. 00:10:57.600 --> 00:11:00.240 I mean, transparency about what the system considered. 00:11:00.639 --> 00:11:04.319 Um, and and this is the one people resist deliberate and efficiency, right? 00:11:04.399 --> 00:11:13.120 You keep uh humans uh doing some work by hand, even when the machine could do it, because you need the skill to stay alive in the org. 00:11:13.439 --> 00:11:20.959 So I think that's the part that makes it responsible deployment and keeps your people sharp instead of quietly hollowing them out. 00:11:21.279 --> 00:11:32.240 So that's interesting, that last point that you brought up, because uh most people, you know, I think the the the average non-technical business professional who was using the tool isn't thinking about that way. 00:11:32.319 --> 00:11:43.439 Like I still need to keep my they they understand human in the loop and those sorts of things, but they don't necessarily understand that for their own benefit, they need to continue to get those repetitions in. 00:11:43.759 --> 00:11:51.919 Like you said, the AI could do it, but I, as in your position, I as the human should take some of that back from AI and continue to do it manually. 00:11:52.240 --> 00:12:08.959 What's the what's the takeaway or the win for the the the power user, let's say, who's really got you know maybe multiple screens going and that sort of thing, and they have started to defer more to the model and they've lost that that that pilot going from autopilot back into manual controls. 00:12:09.360 --> 00:12:15.120 Yeah, the the win is uh that you stay dangerous, and I mean that in the best sense, right? 00:12:15.199 --> 00:12:24.799 So here's what happens to the power user who's running six stabs, three models, and has basically outsourced their first draft uh of thinking to the system. 00:12:25.039 --> 00:12:27.120 I mean, they start to feel incredibly productive. 00:12:27.279 --> 00:12:32.079 I mean, output goes up, uh turnaround shrinks, and the dashboards look great. 00:12:32.159 --> 00:12:34.959 Um, but something quieter is happening underneath, right? 00:12:35.039 --> 00:12:41.919 The small friction that used to make them curious, the moment where you uh go, hey, wait, that doesn't sit right. 00:12:42.159 --> 00:12:49.039 The moment where you rewrite your own sentence because you don't quite believe it yet, those moments get smoothened over, right? 00:12:49.199 --> 00:12:53.919 And so the copilot fills the gap before the gap can do its work on you. 00:12:54.079 --> 00:12:56.480 And the gap is where judgment gets built. 00:12:56.559 --> 00:13:00.559 Like judgment's uh is an accumulated in relationship with consequences. 00:13:00.720 --> 00:13:04.559 Um you get it by being wrong, noticing you were wrong, and updating. 00:13:04.639 --> 00:13:11.279 If the system's always catching you before you fall, you never develop the muscle that uh that tells you the floor is about to give. 00:13:11.519 --> 00:13:14.879 So the practical takeaway for the power user is this, right? 00:13:14.959 --> 00:13:18.079 So pick the decisions in your week that actually matter. 00:13:18.319 --> 00:13:24.559 The ones uh where being caught uh wrong costs you something real and deliberately do those manually first, right? 00:13:24.720 --> 00:13:30.319 Draft your own take before you ask the model, write your own hypothesis before you pull the data. 00:13:30.480 --> 00:13:33.839 I mean, argue with yourself a bit before you argue with the co-pilot. 00:13:34.159 --> 00:13:38.639 And then bring the AI in as a challenger, uh, not as a starter. 00:13:38.799 --> 00:13:42.240 I mean, ask it what evidence would would contradict what you just wrote. 00:13:42.399 --> 00:13:44.319 Ask it for alternative you reject it. 00:13:44.639 --> 00:13:48.240 So I think the other habit I push is uncertainty hygiene. 00:13:48.399 --> 00:13:56.720 When the model gives you an answer, uh force yourself to say out loud how confident you actually are, independent of how confident it actually sounds, right? 00:13:56.799 --> 00:13:59.519 So

confidence in these systems is stylistic. 00:13:59.600 --> 00:14:01.039 I mean, it's not calibrated. 00:14:01.120 --> 00:14:05.199 If you don't check your own gauge, you'll just simply inherit theirs, right? 00:14:05.360 --> 00:14:11.199 So I think the pilots who survived the automation crisis are the ones who still hand flew the plane on clear days. 00:14:11.439 --> 00:14:12.399 Same principle, right? 00:14:12.480 --> 00:14:15.039 I mean, to do the manual reps when the stakes are low. 00:14:15.120 --> 00:14:17.919 So the instinct's there when the autopilot drops out. 00:14:18.240 --> 00:14:20.240 I think that's fantastic advice for the listeners. 00:14:20.319 --> 00:14:24.480 Like, don't fall into the trap of more is better necessarily, right? 00:14:24.720 --> 00:14:33.679 Now, you said that organizations should stop treating AI as a separate capability and start treating it as part of the team's cognitive architecture. 00:14:33.839 --> 00:14:39.279 So what does an org chart look like when a leader starts to uh apply that? 00:14:39.439 --> 00:14:41.439 Like, where does the where do the roles change? 00:14:41.519 --> 00:14:42.879 How do the role definitions change? 00:14:43.039 --> 00:14:46.799 What roles need to get added to an org chart in this environment? 00:14:47.199 --> 00:14:51.839 The org chart stops being a map of humans and starts being a map of decisions. 00:14:52.000 --> 00:14:53.279 I think that's the shift, right? 00:14:53.439 --> 00:14:57.039 So today most org charts answer the question who reports to whom? 00:14:57.199 --> 00:15:03.199 And when AI is generally part of your cognitive architecture, the chart has to also answer who owns the decision. 00:15:03.279 --> 00:15:07.759 I mean, what tools sit inside it and what rules governs how they interact. 00:15:07.919 --> 00:15:09.839 The unit isn't the person anymore. 00:15:09.919 --> 00:15:13.439 I mean, the unit is people plus tools plus rules, right? 00:15:13.600 --> 00:15:20.559 I mean, working on a specific class of decision, pricing decisions, hiring decisions, diagnostic decisions, underwriting decisions. 00:15:20.639 --> 00:15:24.799 I mean, each one is a little cognitive system and needs to be drawn as one. 00:15:25.039 --> 00:15:30.240 So practically a few roles change and a few roles uh show up that weren't there before. 00:15:30.399 --> 00:15:38.159 I mean, your your senior operators, your VPs, your directors, I mean, their job shifts from being the smartest person in the room to being the designer of the room. 00:15:38.319 --> 00:15:45.120 They're deciding which decisions get AI in the loop, at what depth, you know, what with what veto rights. 00:15:45.360 --> 00:15:48.159 Um, that's a design job, not a management job. 00:15:48.319 --> 00:15:50.319 And most of them have never been trained for it. 00:15:50.399 --> 00:15:51.600 So that's the problem, right? 00:15:51.759 --> 00:15:54.399 And then you need roles that didn't exist five years ago, right? 00:15:54.559 --> 00:16:07.279 Someone who owns model behavior and production, not just model performance, someone who owns what the system is allowed to remember about your people and your customers and what it's required to forget. 00:16:07.439 --> 00:16:16.319 Someone who runs what I call the renewal cycle, because any human AI arrangement should have an inspection date and a right to withdraw built into it. 00:16:16.480 --> 00:16:17.519 I mean, that's a real function. 00:16:17.679 --> 00:16:18.639 Call it whatever you want. 00:16:19.199 --> 00:17:13.200 You know, cognitive operations, uh, AI stewardship. 00:17:13.359 --> 00:17:20.079 The name matters less uh than the fact that a human being wakes up on Monday um responsible for it. 00:17:20.240 --> 00:17:22.000 And I think you retire something too. 00:17:22.079 --> 00:17:27.599 I mean, you retire the fiction that the tool is just a tool and therefore doesn't need a seat on the chart. 00:17:27.680 --> 00:17:41.440 I mean, if a system inside your loop of attention and uh judgment uh is actually actually inside your loop of attention judgment, and it's shaping what your team believes before they finish forming the belief, it belongs on the map. 00:17:41.519 --> 00:17:47.440 I mean, leaving it off the chart doesn't make it less powerful, it just makes it unaccountable. 00:17:48.000 --> 00:17:52.240 So you mentioned something and in in the interviews that I've been doing, I've started to see a pattern. 00:17:52.400 --> 00:18:05.920 Uh recently I spoke to somebody and they said that at the individual level, the person is a doer, but in kind of the environment that we're entering into where AI is becoming a generative AI is becoming much more a part of everybody's role, that they need to go from a doer to a planner. 00:18:06.079 --> 00:18:17.119 And that's similar to this concept that you just uh described where uh rather than me being the smartest person in the room as an executive or a leader of a team, I need to now be the designer. 00:18:17.359 --> 00:18:19.440 How does somebody develop that skill? 00:18:19.680 --> 00:18:20.559 Where do you learn that? 00:18:20.639 --> 00:18:22.799 Because you're right, it's not something that's been part of the role. 00:18:23.119 --> 00:18:33.759 Yeah, I mean, I think you'll learn it by running small, cheap experiments on decisions you already own and treating the design of the decision as the artifact, not the outcome. 00:18:34.000 --> 00:18:35.759 Um so here's what I mean, right? 00:18:35.839 --> 00:18:46.960 So pick one recurring decision of uh on your team, not a giant one, something you you would do weekly, a hiring screen, a pricing exception, a content approval, something like that, right? 00:18:47.039 --> 00:18:50.960 So write down on one page how that decision gets made today. 00:18:51.119 --> 00:18:56.319 Who sees what, in what order, with what evidence, and where the judgment call actually happens. 00:18:56.399 --> 00:19:01.599 I mean, most people have never done this

even um you know, for even one of their decisions. 00:19:01.759 --> 00:19:05.279 I mean, when you write it down, you immediately see the scenes. 00:19:05.439 --> 00:19:14.000 I mean, you see where a model would draft, where a human has to weigh, where the veto sits, uh, where the audit trail is missing. 00:19:14.319 --> 00:19:16.480 And that one page is the muscle. 00:19:16.559 --> 00:19:21.359 Uh, do it for 10 decisions, and and you're sort of like a designer, right? 00:19:21.519 --> 00:19:29.119 So the second thing is that this is the part people skip is you have to get fluent enough with the tools uh to have taste. 00:19:29.359 --> 00:19:31.199 Not to build them, to have taste. 00:19:31.599 --> 00:19:39.199 You've never sat with Claude or GPT or Gemini for a weekend and try to get it to do something real inside your domain. 00:19:39.359 --> 00:19:43.599 Um, you know, you cannot design uh a system around it. 00:19:43.759 --> 00:19:47.839 You you'll either overtrust it or undertrust it, and both are expensive. 00:19:48.000 --> 00:20:01.439 So I tell executives uh spend 24, uh 24 hours, you know, uh not 24 hours of demos, um, 24 hours of you alone trying to make the thing useful on a problem you actually care about. 00:20:01.599 --> 00:20:11.599 You'll come out the other side with the physical sense of where it's sharp and where it hallucinates, and what um, and that sense is what lets you draw the room. 00:20:11.759 --> 00:20:15.119 So the third the third piece is you study people who already do this, right? 00:20:15.199 --> 00:20:21.839 So product managers uh at good software companies have been designing human plus system uh workflows for 20 years. 00:20:22.079 --> 00:20:25.039 Air traffic controllers, ICU teams, trading desks. 00:20:25.199 --> 00:20:29.119 Go read how they think about handoffs, alarms, override authority. 00:20:29.199 --> 00:20:30.559 I mean, that literature exists. 00:20:30.720 --> 00:20:34.480 I mean, most business leaders have never touched it because it wasn't their job. 00:20:34.639 --> 00:20:35.439 Now it is. 00:20:35.679 --> 00:20:39.199 You know, so nobody's going to hand you a certification for this. 00:20:39.279 --> 00:20:45.119 You build it by doing it on something small in public with your own team watching. 00:20:45.439 --> 00:20:46.879 Yeah, that's fantastic advice. 00:20:46.960 --> 00:20:54.559 And for the listener, one of the ways that we've kind of seen it at Chief AI Officer is that before we would certify a chief AI officer, there was something that I was looking for. 00:20:54.639 --> 00:20:57.279 And it wasn't that they were using AI like a drive-thru window. 00:20:57.359 --> 00:20:59.279 Hey, I need an answer, got it, I'm out. 00:20:59.519 --> 00:21:24.079 It was that individual who spent, as Rana's suggesting, an entire weekend in one thread, really going deep and paying attention to stuff and seeing how like, and if you have never done that as a listener, I would encourage you to rather than just bounce

between threads, go deep and see the difference, and you will understand the model behavior for your role and your use in a way that you never have before. 00:21:24.319 --> 00:21:29.519 So, one of the concepts, Rana, that that's talked about in the book is this idea of attention sculpting. 00:21:29.919 --> 00:21:31.519 Can you tell me what you mean by that? 00:21:31.919 --> 00:21:41.519 Yeah, attention sculpting is what happens uh when a system decides on your behalf and usually invisibly what you notice, right? 00:21:41.599 --> 00:21:48.399 And because what you notice becomes what you value, the thing shaping your attention is quietly shaping your preferences. 00:21:48.639 --> 00:21:50.240 So think about the mechanics, right? 00:21:50.319 --> 00:21:59.439 So every feed you touch, every recommender, every autocomplete, um, every rank search result is making a claim about salience, right? 00:21:59.519 --> 00:22:08.240 It's saying out of the ocean of things you could be looking at right now, um, these are the ones worth your milliseconds. 00:22:08.319 --> 00:22:11.919 I mean, you experience uh you experience that as a convenience. 00:22:12.079 --> 00:22:15.439 What's actually happening is curation with a point of view. 00:22:15.599 --> 00:22:23.199 The system has a model of view and it's pulling on the levers of what surfaces and in what order with what emotional temperature. 00:22:23.359 --> 00:22:34.240 The the reason I call it sculpting and not uh say filtering is that filtering sounds subtractive, like like a like the system just removes noise. 00:22:34.559 --> 00:22:37.199 Sculpting is additive and it's shaping, right? 00:22:37.279 --> 00:22:41.919 A sculptor doesn't just take marble away, they reveal a specific form. 00:22:42.079 --> 00:22:53.279 I mean, these systems are revealing a specific version of the world to you over and over, and the version they uh reveal trains what your attention learns to reach for the next time. 00:22:53.439 --> 00:22:56.079 You know, TikTok is the crude, obvious example, right? 00:22:56.240 --> 00:23:08.559 But the same mechanic runs inside uh a Gmail priority inbox, inside a code assistant suggestion ranking, inside which three documents your enterprises search puts on top, right? 00:23:08.639 --> 00:23:12.720 So now pair that with what you uh you were just describing, right? 00:23:12.799 --> 00:23:14.720 The person who spends the weekend in one thread. 00:23:14.879 --> 00:23:17.679 That person is doing the opposite of being sculpted. 00:23:18.159 --> 00:23:22.399 They're pushing back on the system's sense of uh salience and imposing their own. 00:23:22.559 --> 00:23:25.919 I mean, they're saying, I decide what deserves that PA. 00:23:26.399 --> 00:23:26.639 Right? 00:23:26.720 --> 00:23:27.599 Not the interface. 00:23:27.759 --> 00:23:31.439 That's the muscle that atrophies when you live in the drive-thru mode. 00:23:31.599 --> 00:23:35.919 I mean, you stop noticing that something else is doing the noticing for you. 00:23:36.159 --> 00:23:43.119 So where this gets uh, I guess, ethically loaded is that attention sculpting doesn't uh ask for consent in any meaningful way. 00:23:43.279 --> 00:23:48.639 There's no moment where the system says, Hey, I'm about to ship what feels important to you for the next four hours. 00:23:48.720 --> 00:23:49.359 Is that okay? 00:23:49.519 --> 00:23:49.759 Right? 00:23:49.919 --> 00:23:50.639 It just happens. 00:23:50.799 --> 00:24:00.240 And in a hybrid setup, where the assistant finishes your sentences and uh mirrors your tone, the the sculpting stops feeling external. 00:24:00.399 --> 00:24:02.639 It just feels like you had the thought. 00:24:02.799 --> 00:24:06.000 That's the part leader should be losing sleep over. 00:24:06.799 --> 00:24:14.399 Now, in that in that vein, I guess, uh, there's this self-check of am I still deciding or am I just approving? 00:24:14.720 --> 00:24:18.159 So how would you suggest a leader kind of audit? 00:24:18.240 --> 00:24:21.679 I mean, maybe as an individual, you can stop and you can ask yourself those questions. 00:24:21.759 --> 00:24:26.000 Maybe you've got a sticky note that says, you know, am I deciding or am I just approving? 00:24:26.159 --> 00:24:30.000 But how do you audit that in the team's behavior and their use of the tools? 00:24:30.559 --> 00:24:35.679 The tell isn't in the decision, it's the it's in the artifacts around the decision, right? 00:24:35.759 --> 00:24:42.079 So if you want to audit uh whether your team is still deciding or just approving, don't ask them. 00:24:42.159 --> 00:24:44.559 I mean, uh they'll tell you they're deciding. 00:24:44.720 --> 00:24:46.240 Everyone thinks they're deciding. 00:24:46.720 --> 00:24:48.399 Look at what they produce, right? 00:24:48.480 --> 00:24:50.399 The three things that actually inspect, right? 00:24:50.480 --> 00:24:52.000 First, the reasoning trail. 00:24:52.159 --> 00:25:03.119 Um, when someone on your team green lights and AI assisted recommendation, is there a written record of why they agreed and specifically what they considered and rejected? 00:25:03.199 --> 00:25:06.480 Uh, not a summary of the model's output, their reasoning. 00:25:06.639 --> 00:25:12.720 I mean, if if if the doc reads like a restatement of what the system suggested, they approved. 00:25:12.879 --> 00:25:18.960 If it reads like the system said A, I checked B and C, and here's why A still wins, they decide it. 00:25:19.119 --> 00:25:22.319 You can spot the difference in about 30 seconds per artifact. 00:25:22.480 --> 00:25:22.639 Right? 00:25:22.879 --> 00:25:26.319 I mean, the second disagreement uh rate, right? 00:25:26.399 --> 00:25:29.439 And so track how often your team overrides the tools. 00:25:29.599 --> 00:25:35.359 I mean, if if a team is running a copilot for three months and the override rate is under 5%, something is wrong. 00:25:35.519 --> 00:25:39.439 I mean, either the tool is superhuman, which it isn't, or the humans have gone

quiet. 00:25:39.599 --> 00:25:41.919 I mean, healthy hybrid work has friction in it. 00:25:42.000 --> 00:25:46.799 You want to see people pushing back, asking for alternatives, requesting the counterfactual. 00:25:46.879 --> 00:25:49.279 If everyone's nodding, uh nobody's thinking. 00:25:49.599 --> 00:25:49.759 Sure. 00:25:50.000 --> 00:25:56.799 And I think the third, and I think it's the one most leaders won't do is run a periodic unplugged decision. 00:25:56.879 --> 00:26:02.240 Like pick a recurring call once a month, one person on the team makes it without the AI in the loop. 00:26:02.319 --> 00:26:03.199 Then compare. 00:26:03.359 --> 00:26:06.480 Not to prove the human is better, but to see whether the human still can. 00:26:06.559 --> 00:26:13.519 I mean, the moment someone says, I don't know how I'd even start without a tool, you've found your uh you found your de-skilling. 00:26:13.599 --> 00:26:14.399 I mean, that's just signal. 00:26:14.559 --> 00:26:14.799 Yeah, yeah. 00:26:15.039 --> 00:26:21.839 So the meta movement underneath all three is that you're you're building an audit trail of human cognition, not of AI output. 00:26:22.079 --> 00:26:28.079 Most companies log what the models did, but almost nobody logs what the human thought flip that, right? 00:26:28.159 --> 00:26:31.119 So the question isn't what the system would recommend. 00:26:31.199 --> 00:26:36.000 Uh the question is what did you weigh and what would have changed your mind? 00:26:36.159 --> 00:26:39.919 And if your team can't answer that on demand, um, they're just approving. 00:26:40.399 --> 00:26:43.359 So, you know, uh I guess there's a strata of usage. 00:26:43.519 --> 00:26:50.960 Some might be, hey, I need the email rewritten, some might be I need to do strategic action for the company and I need I need the support. 00:26:52.079 --> 00:27:10.799 That approach is sound, however, it seems like at some point on that strata of I guess uh difficulty or complexity of the use of the model, it would be cumbersome at the low end of the strata and extremely valuable at the high end of the strata. 00:27:10.960 --> 00:27:23.519 How does somebody distinguish, like, when should I really get this serious about it and try my decision making and be prepared to defend the reason why I presented this artifact as compared to something else? 00:27:24.319 --> 00:27:28.000 Yeah, I think the threshold is reversibility. 00:27:28.159 --> 00:27:32.879 That's the single question that tells you where you are on the strata, right? 00:27:33.039 --> 00:27:41.759 If the thing you're about to ship can be undone cheaply in an hour with an apology and a resent, you don't need the whole reasoning trail. 00:27:42.000 --> 00:27:44.240 Um, you know, rewrite the email, move on, right? 00:27:44.399 --> 00:27:46.639 Life is short and cognitive overhead is real. 00:27:46.799 --> 00:27:56.720 So if if you audit every keystroke, you'll grind to a halt and your team will start hiding their AIUs from you, which is worse than any of the harms you were trying to even prevent, right? 00:27:56.960 --> 00:28:03.439 So so the moment uh but the moment uh decision becomes hard to reverse, the calculus flips, right? 00:28:03.599 --> 00:28:06.879 Hard to reverse means it changes what somebody else does next. 00:28:07.119 --> 00:28:21.519 A pricing move, competitors will match, a hire, a public commitment, um, you know, a model going into production against real customers, um, the a strategy dock that becomes the frame everyone else plans against for the next two quarters. 00:28:21.679 --> 00:28:28.000 I think those are the decisions where you owe yourself uh and the people downstream of you are a record of why this and not that. 00:28:28.240 --> 00:28:34.159 So I I kind of, you know, I use a rough um three-band test in my own work. 00:28:34.240 --> 00:28:40.000 Like band one, um reversible in under a day, no artifact needed, just ship. 00:28:40.159 --> 00:28:46.639 Band two, reversible but expensive, meaning it'll cost you a week or a relationship to unwind. 00:28:46.799 --> 00:28:51.759 You write two lines, what the system suggested, what you actually chose, and why they differ. 00:28:51.919 --> 00:28:52.559 Two lines. 00:28:52.639 --> 00:28:53.039 That's it. 00:28:53.199 --> 00:29:01.119 You know, uh band three, functionally irreversible or high stakes for someone else's life or money, you write the contrastive version. 00:29:01.279 --> 00:29:08.000 I mean, I chose A or B because of X and Y, even though B offered Z, and you keep it, right? 00:29:08.240 --> 00:29:15.519 Um the other trigger uh separate from reversibility is um asymmetry of harm. 00:29:15.599 --> 00:29:25.679 Um if the downside lands on someone or somebody who wasn't in the room when you decided, um, that's a band three, no matter how small the decision feels to you. 00:29:25.839 --> 00:29:34.079 Like, you know, a collection script, a screening filter, a routing rule, uh small to you, but enormous to the person on the other end. 00:29:34.240 --> 00:29:48.079 So the honest answer to when it gets serious is like, you know, uh when future you or a regulator or a person uh affected would reasonably ask you to show your work, um, you know, that's when. 00:29:48.399 --> 00:29:51.679 So if you show it now, it's cheaper than reconstructing it later. 00:29:52.079 --> 00:29:52.639 Heck yeah. 00:29:52.720 --> 00:29:54.079 Okay, that's very, very helpful. 00:29:54.240 --> 00:29:58.960 Now I'm gonna shift gears a little bit because I I want to kind of dig into how you actually use it. 00:29:59.119 --> 00:30:05.679 So um maybe take a second and walk me through what you do before you open the model on a strategic decision. 00:30:05.759 --> 00:30:07.279 Like what would you write down? 00:30:07.359 --> 00:30:10.159 How how long would that process take? 00:30:10.559 --> 00:30:18.720 Yeah, I mean, um before I open the model, I write three things down on paper actual paper or a blank doc with nothing else on

the screen, right? 00:30:18.799 --> 00:30:22.799 This takes me about 10 minutes, and it's the most important 10 minutes of the whole session, right? 00:30:22.960 --> 00:30:28.879 First thing I'd write, what am I actually trying to decide in one sentence with the decision maker named? 00:30:29.039 --> 00:30:30.399 Not think about pricing. 00:30:30.559 --> 00:30:38.639 Something like I need to decide by Friday whether we raise enterprise pricing 15% for new logos only, and I'm the one signing off. 00:30:38.799 --> 00:30:41.679 If I can write that sentence, I'm not ready to use the tool. 00:30:41.759 --> 00:30:42.960 I'm ready to journal. 00:30:43.119 --> 00:30:44.559 Those are two different activities. 00:30:44.799 --> 00:30:45.119 Right? 00:30:45.279 --> 00:30:48.000 And the second thing, what's my current instinct and why? 00:30:48.159 --> 00:30:48.319 Right? 00:30:48.399 --> 00:30:49.359 Two or three sentences. 00:30:49.439 --> 00:30:56.559 Like this matters because once the model gives me a fluent, well-structured answer, my prior gets quietly overridden. 00:30:56.720 --> 00:30:59.439 I won't remember what I actually thought coming in. 00:30:59.599 --> 00:31:00.399 So I pin it. 00:31:00.480 --> 00:31:01.679 Later I can compare. 00:31:01.919 --> 00:31:08.879 And um, if I've drifted, I want to know whether I drifted because of a real argument or because the output just sounded confident. 00:31:09.119 --> 00:31:09.279 Right? 00:31:09.439 --> 00:31:11.359 Third, what would change my mind? 00:31:11.519 --> 00:31:15.199 This is the one people skip, and it's the one that does the most work. 00:31:15.359 --> 00:31:21.279 I write down two or three specific pieces of evidence and arguments that, if true, would flip my instinct, right? 00:31:21.439 --> 00:31:26.879 Churn data over a threshold, a competitor move, whatever that is, like a customer segment I hadn't considered. 00:31:27.039 --> 00:31:28.159 Now I have a test. 00:31:28.399 --> 00:31:33.839 Um, now I have a test the model has to pass instead of a why but it has to match, right? 00:31:33.919 --> 00:31:37.759 Um so then you open the model and I don't ask it what to do. 00:31:37.919 --> 00:31:48.240 Um, you know, I ask it to argue against my instinct using a criteria I just wrote, and I asked it to um find the evidence I said would change my mind and tell me honestly if it exists. 00:31:48.319 --> 00:31:51.519 I mean, I ask it to name what I'm probably not seeing. 00:31:51.679 --> 00:31:56.000 The whole pre-work is maybe 10 minutes, the session after is 30, 40. 00:31:56.240 --> 00:32:02.000 And the reason the 10 matter so much is that it makes me a client with a brief, not a user with a prompt. 00:32:02.159 --> 00:32:06.960 You know, the system's now working on my judgment instead of substituting it, and that's the whole game. 00:32:07.279 --> 00:32:07.919 I love it. 00:32:08.079 --> 00:32:14.879 So that that really feeds into what you were talking about earlier, like forcing the repetitions of the human still participating in the process. 00:32:15.439 --> 00:32:22.319 And listeners, if you're if you were to follow that process, I think it becomes obvious that like you are very involved. 00:32:22.480 --> 00:32:32.799 You're not just absorbing output, you're now actively participating as a partner in the thinking rather than outsourcing that thinking to, hey, what should I do? 00:32:33.039 --> 00:32:34.319 Open-ended kind of question. 00:32:34.559 --> 00:32:37.199 What would be some, you know, we we talked about this. 00:32:37.519 --> 00:32:57.039 One of the concerns that we we hear across a bunch of industries is that particularly like legal and things where the years of being in that role allows you to be the person who's now capable of training that junior, training the person that's gonna re that you can delegate to or that may eventually replace you as you you move on or or ascend in the organization. 00:32:57.279 --> 00:33:12.240 In 10 years, where are these or five years, where are these senior people gonna come from if if AI is doing a lot of the heavy lifting, even though we're participating, as you just suggested, you know, the the pre-work before we go into the models? 00:33:13.119 --> 00:33:27.919 How do people with that concern, how do they build something today that makes sure that they're not gumming down the human as AI gets more capable and becomes more of a part of their operational modus operandi, I guess? 00:33:28.240 --> 00:33:28.879 Yeah. 00:33:29.359 --> 00:33:35.039 Um, you know, replacement is a story we tell uh to make budget decisions easier, right? 00:33:35.199 --> 00:33:39.839 Augmentation is what actually happens on like a Tuesday afternoon, right? 00:33:40.079 --> 00:33:43.119 So I give you the frame I actually use, right? 00:33:43.199 --> 00:33:52.639 So every job is a bundle of tasks, and every task sits uh somewhere on a spectrum from pure pattern recognition to pure judgment under ambiguity, right? 00:33:52.720 --> 00:33:59.599 So AI is coming for the pattern recognition and first, you know, hard and it's already there in a lot of places. 00:33:59.759 --> 00:34:02.079 Uh look at what happened in customer support. 00:34:02.159 --> 00:34:08.159 Like Klarna said in early 2024, their AI assistant was doing the work of 700 agents. 00:34:08.240 --> 00:34:08.800 That's real. 00:34:08.960 --> 00:34:16.480 But if you actually look at what stayed, it's the escalations, the weird ones, the calls where somebody's angry and the policy doesn't quite fit. 00:34:16.559 --> 00:34:21.360 You know, those got harder, not easier, because the easy ones stop landing in human cues. 00:34:21.519 --> 00:34:24.880 So the line I draw isn't augmentation versus replacement. 00:34:25.039 --> 00:34:29.840 It's which tasks in this role are compressible and what's left when you compress them. 00:34:29.920 --> 00:34:34.079 And then critically, is the leftover job actually a job a human wants to do? 00:34:34.240 --> 00:34:39.280 Or have

you built a role that that's nothing but exception handling and blame absorption? 00:34:39.679 --> 00:34:43.760 You know, and I think the last part is where I think most executives are getting it wrong right now. 00:34:43.920 --> 00:34:48.800 They're modeling headcount reduction and calling it augmentation because it sounds gentler. 00:34:49.039 --> 00:35:01.920 But if you strip the routine 80% out of the out of a role and leave a person doing nothing but the hard 20, you created a job with no on-ramp, no recovery time, and no way to build intuition. 00:35:02.080 --> 00:35:06.320 I mean, junior people can't grow into it uh because the practice reps are gone. 00:35:06.480 --> 00:35:09.360 And senior people burn out because there's no rhythm, right? 00:35:09.440 --> 00:35:10.880 There's that's not augmentation. 00:35:10.960 --> 00:35:16.800 That's that's a role that will quietly fail in 18 months, and you will blame the person, right? 00:35:17.040 --> 00:35:28.320 So the honest version of augmentation is when the human's decision quality goes up and the human is still doing enough uh of the work, it stays sharp uh to stay sharp at it. 00:35:28.480 --> 00:35:36.400 You know, uh radiologists reading with the eye assist and catching more, uh, not radiologists rubber stamping a queue of 400 scans and a shift. 00:35:36.560 --> 00:35:41.600 I mean, those are different features, and we're choosing between them right now, mostly without admitting we're actually choosing. 00:35:42.080 --> 00:35:45.360 So as we as you come to the end of this, this has been awesome. 00:35:45.440 --> 00:35:46.720 I'm looking forward to getting the book. 00:35:46.800 --> 00:36:02.000 Uh, and we're gonna have, for the listener, we're gonna have links to uh order the book uh right there on the show notes, and I would encourage you to do it because I think that if nothing else, maybe you've you've watched a lot of tactical trainings on how to use the tools better, how to build a skill, whatever those things might be. 00:36:02.160 --> 00:36:11.520 Maybe you've had uh external experts come in and work with the team to help them start to integrate generative AI into the operations and the workflows and the processes and that sort of thing. 00:36:11.840 --> 00:36:14.960 But if you're listening to this, you are a decision maker, right? 00:36:15.120 --> 00:36:27.600 And I think that if you I mean, I I I'll admit, some of what you've talked about as not necessarily best practice for the human uh AI loop, I've been guilty of for sure. 00:36:27.840 --> 00:36:40.640 Um so as a listener, I would encourage you to practice those three steps, the pre-work that Rana shared earlier, and see if the experience and particularly the quality of the output that you're getting improves. 00:36:40.800 --> 00:36:42.400 I have no doubt that it will. 00:36:42.560 --> 00:36:48.880 Um, and again, I'll admit I've been guilty of not necessarily doing that uh exclusively on the strategic work. 00:36:48.960 --> 00:36:50.400 So that was a big takeaway for me. 00:36:50.560 --> 00:36:56.640 Now let's talk about tomorrow for the listener or Monday morning, you know, if they're listening to this over the weekend. 00:36:56.880 --> 00:37:08.720 Let's say it's a CEO, they they they listen to this, they're running a$50 million company, maybe they've got Chat GPT business deployed for their teams, but there's no real rules like we talked about earlier with human plus tool plus rules. 00:37:09.040 --> 00:37:15.840 Maybe they don't have that audit trail where an individual who's providing output that was synthesized from the models is able to defend it. 00:37:16.160 --> 00:37:25.840 And maybe they don't even have any idea of whether or not they haven't they haven't had a uh an open discussion about this idea of deciding versus approving. 00:37:26.080 --> 00:37:34.800 What would you suggest be one of the first steps that they do to start to introduce this into um how their company behaves when using the models? 00:37:35.280 --> 00:37:48.560 Yeah, so the cleanest test I know, can you point to the exact moment in your workflow where a human has enough context, enough time, and enough authority to say no to the model? 00:37:48.720 --> 00:37:52.800 And does that moment actually happen or is it just theater? 00:37:53.040 --> 00:37:53.360 Right? 00:37:53.600 --> 00:37:57.200 Because most human in the loop designs I see are theater. 00:37:57.360 --> 00:38:08.720 I mean, there's a checkbox, someone clicks approve, but the interface has already framed the answer, the queue has already prioritized speed, um, and the reviewer has like 30 seconds uh and 40 items. 00:38:08.880 --> 00:38:09.920 That's not oversight. 00:38:10.080 --> 00:38:11.600 I mean, that's just laundering, right? 00:38:11.760 --> 00:38:17.520 So the responsible deployment starts by taking that seriously and designing against it. 00:38:17.600 --> 00:38:26.320 Uh, and I a few things I look for when I'm evaluating the system ours or someone else's um are like one, uh calibrated confidence, right? 00:38:26.400 --> 00:38:32.960 So the model needs to tell you how sure it is, and that number needs to actually mean something across the distribution. 00:38:33.200 --> 00:38:39.200 If it says 90% and it's write 60, you've built a machine that manufactures false trust. 00:38:39.360 --> 00:38:42.000 You know, calibration isn't a nice to have, right? 00:38:42.080 --> 00:38:49.440 It's it's a it's the substrate uh under every downstream decision a human makes uh uh about whether to defer or push back. 00:38:49.920 --> 00:38:56.400 Two, um, audit trails that uh that a non-engineer can read, not logs, trails. 00:38:56.800 --> 00:39:01.440 Um the what, the why, the inputs, the alternatives that were considered and rejected. 00:39:01.600 --> 00:39:07.680 Because when something goes wrong, and it will, I mean, you need to be able to reconstruct the decision without a forensic team. 00:39:07.760 --> 00:39:09.920 And if you can't,

you don't really have accountability. 00:39:10.080 --> 00:39:11.600 You have an accountability gap. 00:39:11.760 --> 00:39:21.040 Um that's yeah, that AI, uh the AI is a black box, um, and and the operators uh you know says they were just following it, and nobody's actually responsible. 00:39:21.120 --> 00:39:23.520 Um and then it becomes a failure mode. 00:39:23.680 --> 00:39:29.760 Um three, watch for what I call the moral crumple zone, which we discussed earlier. 00:39:29.920 --> 00:39:35.920 When the systems fail, blame tends to land on the human operator who had the least real control. 00:39:36.240 --> 00:39:39.600 The clinician who accepted the flag, the analyst who signed off. 00:39:39.760 --> 00:39:44.560 So if your deployment design lets that happen, you haven't deployed responsibly. 00:39:44.720 --> 00:39:48.000 You've built a liability sheet with the person shape dent in it. 00:39:48.240 --> 00:39:52.160 So the last one, the fourth, plan for de-skilling. 00:39:52.320 --> 00:39:53.680 You just have to plan for it. 00:39:53.840 --> 00:40:00.800 I mean, if the human override exists on paper, but the humans never actually override in 18 months, they've lost a skill too. 00:40:00.960 --> 00:40:06.720 You know, uh we talked about aviation, which figured out decades ago and it called it automation complacency. 00:40:06.880 --> 00:40:09.120 You have you have to force the muscle to stay alive. 00:40:09.280 --> 00:40:16.480 I mean, structured disagreement, um, rotation cases where the humans decide without seeing the model first and then compares. 00:40:16.800 --> 00:40:18.400 Um, those things become really important. 00:40:18.560 --> 00:40:20.800 Otherwise, the loop is closed and you just haven't noticed. 00:40:20.960 --> 00:40:24.720 So I think, you know, responsible isn't like a certification. 00:40:24.800 --> 00:40:31.200 It's whether any of that um uh any of that is actually happening on a Wednesday afternoon when nobody's watching. 00:40:31.600 --> 00:40:32.080 Wow. 00:40:32.240 --> 00:40:36.640 Um a whole different way of thinking about AI, dear listener, for sure. 00:40:36.720 --> 00:40:47.680 And going from, oh, now I can move faster to wait a minute, maybe I need to slow down for a second on some of these activities to make sure that the output that I'm getting isn't just slop, as they call it, right? 00:40:47.920 --> 00:40:50.320 So, Rana, final question. 00:40:50.640 --> 00:40:56.960 What is a question that you wish more executives or or potential readers would ask you? 00:40:57.120 --> 00:40:58.960 And what would be your answer to that? 00:40:59.600 --> 00:41:04.960 Every founder I talk to uh can tell me what AI is doing uh for their output. 00:41:05.040 --> 00:41:09.280 You know, faster tickets, uh cheaper content, better code review, great. 00:41:09.520 --> 00:41:15.600 What almost none of them can tell me is what six months of using these tools have done to how their people think. 00:41:15.680 --> 00:41:33.600 You know, whether their designers uh still sketch before prompting, whether the PMs still write the spec themselves or just edit the one the model gave them, whether the their engineers can still read a stack trace cold, or whether they've quietly become dependent uh on pasting it into um the chat window. 00:41:33.920 --> 00:41:36.960 So I think that's where the actual risk lives. 00:41:37.040 --> 00:41:39.200 I mean, it's not in the tool, it's in the drift. 00:41:39.360 --> 00:41:45.520 Um so and here's why I'd give this advice over uh above everything, anything else. 00:41:45.600 --> 00:41:55.200 As a founder, you have maybe um 18 months to build a team whose judgment you can trust when things get weird, and and and things always get weird. 00:41:55.360 --> 00:42:02.000 You know, the fundraiser blows up and the customer turns and the the model, the provider uh changes pricing overnight. 00:42:02.080 --> 00:42:08.080 Um in those moments, you need people who can reason from first principles, not people who can prompt well. 00:42:08.160 --> 00:42:20.720 And if you spend your first year optimizing for velocity through AI, you may find you you've built a company that can execute beautifully on well-defined tasks and falls apart the moment task isn't well defined. 00:42:20.880 --> 00:42:21.120 Right? 00:42:21.280 --> 00:42:32.080 So I think um practical version, I guess, you know, once a quarter, pick a real problem in your business and have your team solve it without any AI assistance, not as a hazing ritual. 00:42:32.320 --> 00:42:32.720 Yeah. 00:42:33.040 --> 00:42:34.720 You know, as a diagnostic, right? 00:42:34.880 --> 00:42:36.000 Watch what happens. 00:42:36.240 --> 00:42:47.440 And if a room goes quiet, if people can't get started, um, if the quality falls off a cliff, you've learned something important about what you've actually been building, and you still have time to correct it, hopefully, right? 00:42:47.600 --> 00:42:53.120 The founders who win the next decade won't be the ones who um adopted AI fastest. 00:42:53.280 --> 00:43:00.720 They'll be the ones who noticed early what it was doing to the minds of the people around them and made deliberate choices about it. 00:43:01.280 --> 00:43:01.920 Wow. 00:43:02.320 --> 00:43:07.920 That's a very interesting exercise, and I think I might present that to some of our clients that we've worked with for you know longer than six months. 00:43:08.000 --> 00:43:08.560 It's fantastic. 00:43:08.720 --> 00:43:10.560 So, Ron, again, congratulations on the book. 00:43:10.640 --> 00:43:12.480 Uh, the book is The AI Instinct. 00:43:12.560 --> 00:43:13.600 It's out by Wiley. 00:43:13.680 --> 00:43:14.400 It's out now. 00:43:14.480 --> 00:43:18.640 You can get it um anywhere that you're you're buying books, including Amazon. 00:43:18.800 --> 00:43:35.120 Um, and a number of takeaways from me personally, um, I'm going to be introducing several of these concepts into the trainings that we're doing for organizations because we have been more focused on the speed, the efficiency, the quality productions that are possible. 00:43:35.280 --> 00:43:37.040 And it's been kind of an open loop for me. 00:43:37.120 --> 00:43:38.320 Well, what does that do to the human? 00:43:38.400 --> 00:43:53.840 And I think that this model that's been uh discussed today and is discussed in detail in the book, obviously, uh is gonna be a new addition to perspectives that we introduce to the leadership as they're uh seriously considering uh AIFI and their team. 00:43:53.920 --> 00:43:54.960 So, Rana, again, thank you. 00:43:55.040 --> 00:43:56.720 I wish you all the success with the book. 00:43:56.880 --> 00:44:03.200 And uh listeners, we're gonna have all of the information related to the book and how to get it, and more information about how to get in touch with Rana in the show notes. 00:44:03.360 --> 00:44:06.480 And uh once again, thank you for uh being a listener of the show. 00:44:06.720 --> 00:44:07.760 I really appreciate it. 00:44:08.000 --> 00:44:19.760 If this episode, if you got some as many takeaways as I did from this, please feel free to forward it to anybody else who's leading a team, who's asking these questions about how do we make sure that we're using AI at work correctly. 00:44:19.920 --> 00:44:22.080 So, Rana, again, thank you so much. 00:44:22.480 --> 00:44:23.280 Thank you, Chris. 00:44:23.360 --> 00:44:24.240 Uh, it was a real pleasure. 00:44:24.400 --> 00:44:25.120 Thank you for having me. 00:44:25.360 --> 00:44:25.840 You're welcome. 00:44:26.000 --> 00:44:28.320 Thanks for tuning in to using AI at work. 00:44:28.720 --> 00:44:32.720 Don't forget to subscribe for more conversations about how to use AI at work. 00:44:33.120 --> 00:44:39.600 And a special thank you to our sponsor, Chief AI Officer, for empowering businesses with AI education and training. 00:44:39.760 --> 00:44:46.640 Visit their website for a free AI readiness assessment and AI strategy guide to help you get started using AI at work. 00:44:46.880 --> 00:44:51.280 That's www.chiefaiofficer.com. 00:44:51.440 --> 00:45:03.360 Follow us on Twitter at the handle UsingAI at work and visit www.usingai at work.com for free resources to help you harness AI in your role.