Latest / Humanity Working / Why is AI making some of us more burned out?

Transcript

Why is AI making some of us more burned out? · Humanity Working

• 0:14
  SPEAKER_03: So, Paul, you have a series coming out now called Seven Big Questions About AI, the very uh rolls off the tongue acronym, seven BQ AAI.
• 0:28
  SPEAKER_01: I think all acronyms should be like that, but completely incomprehensible and not memorable at all.
• 0:34
  SPEAKER_03: So my first question is: how did you figure out only seven questions?
• 0:39
  SPEAKER_03: How did you narrow it to only seven big questions?
• 0:43
  SPEAKER_02: Yeah, well, I will admit that there was sort of like a um a thought that was going on through my head around the fact that we've got the Magnificent Seven.
• 0:53
  SPEAKER_02: Wouldn't it be interesting if there were seven questions associated directly with the Magnificent Seven?
• 0:59
  SPEAKER_02: And it kind of said like the idea started, uh started there.
• 1:03
  SPEAKER_04: Yeah.
• 1:03
  SPEAKER_02: But then I was initially I was like, okay, well, that means we need like a question associated with readiness and a question associated with organization and so on and so forth.
• 1:14
  SPEAKER_02: But it didn't, that didn't really work, it didn't really work.
• 1:17
  SPEAKER_02: So I kind of abandoned that idea like fairly rapidly.
• 1:20
  SPEAKER_04: Yeah.
• 1:21
  SPEAKER_02: And then I just started to think about, okay, well, what are the questions that as I've been doing the virtual book tour and as I've been talking to uh to people uh around this topic, and as I've been doing research for my next book, uh, which uh should be on the next few months, what are the questions that people ask that are either questions that people are asking a lot um you know, lots of people are asking the same question, but actually more, what are the questions whereby when they ask those questions, the answer is, whew, yeah.
• 1:59
  SPEAKER_02: Well, I'm glad you asked that question.
• 2:01
  SPEAKER_02: That's actually really difficult or complicated to answer, where there's sort of an onion that you can peel back.
• 2:06
  SPEAKER_02: And so having abandoned the idea of like mapping to each of the individual layers, I'm just writing those things down, and then magically it ended up a seven.
• 2:19
  SPEAKER_02: So I was like, all right, we'll we'll stick with that.
• 2:22
  SPEAKER_02: So um, but it was interesting.
• 2:24
  SPEAKER_02: I did I I didn't use AI to help me with that, and I'm kind of glad it didn't, because it reminds you of the things that AI can't really do, right?
• 2:35
  SPEAKER_02: AI can't be the equivalent of a question that leads to a one or two hour discussion.
• 2:43
  SPEAKER_02: And it can obviously give you some thoughts, it can stretch your own thinking, it can do all of those kind of things, but it doesn't it can't observe which questions lead to really long discussions about things.
• 2:53
  SPEAKER_02: And so this was these were really the questions that did that.
• 2:57
  SPEAKER_02: And right now there's seven, and who knows, there might be more an eighth or a ninth that pop up in the uh in the not too distant future, but right now it's seven.
• 3:05
  SPEAKER_03: And that beautiful acronym will fade into oblivion.
• 3:08
  SPEAKER_03: And and to your point, it would have been interesting because it would have given you a question per seven.
• 3:14
  SPEAKER_03: It would have matched up those questions, and you might have just gone off, and that would have been it.
• 3:18
  SPEAKER_02: Yeah, that's what that's what AI would have done.
• 3:20
  SPEAKER_02: Yeah, it's interesting to think about that.
• 3:22
  SPEAKER_03: Yeah, so why don't you tell uh the listeners what what is this first big question and what you were thinking when you came up with that question, and then we'll dig into it a little bit.
• 3:32
  SPEAKER_02: Yeah, so the first question, the question that we're gonna spend some time meandering around today, doing our little version of the multi-hour discussions about it, um, is why is AI making some of us more burned out?
• 3:47
  SPEAKER_02: And you'll notice there's some caveats in there and all that kind of thing.
• 3:50
  SPEAKER_02: It's not like why is AI making us burn more burned out?
• 3:53
  SPEAKER_02: It's why is AI making some of us more burned out?
• 3:56
  SPEAKER_02: It's not is AI making some of us burned out, it's why is AI making some of us more burned out.
• 4:01
  SPEAKER_02: So we'll get into some of those uh some of those nuances, I'm sure, as we go.
• 4:05
  SPEAKER_03: Yeah, so so I think this was a really interesting one because um, you know, we're supposed to get time back from AI.
• 4:14
  SPEAKER_03: That's the promise, is where, oh, it's gonna do so many things for you.
• 4:18
  SPEAKER_03: Um anybody that's spent enough time in it in the last you know year or two as it's evolved and they start applying it in in real life.
• 4:26
  SPEAKER_03: Um, I know I've certainly noticed this.
• 4:29
  SPEAKER_03: Um I don't know that it's given us time back.
• 4:32
  SPEAKER_03: It feels more like it's raised the bar in some ways of expectation because you can do things so much easier, but but it's kind of it it's not delivering on that initial promise either.
• 4:50
  SPEAKER_03: Yeah.
• 4:50
  SPEAKER_03: And I think that might be contributing uh quite a bit to to that issue and uh that frustration, you know.
• 4:57
  SPEAKER_03: And we talked about before, and then and then I'll let you speak to this a little bit.
• 5:01
  SPEAKER_03: We talked about before like when you're doing graphics with AI, and you're like every time it's repeating the same error, but it's not correcting the one you're telling it to correct, or it only corrects what you tell it and screws something else up.
• 5:16
  SPEAKER_03: So now I've spent I initially got the the rush of the first image.
• 5:21
  SPEAKER_03: Oh, that's pretty good, and then I spend a half hour tweaking it and getting more and more frustrated, and then I've completely lost the flow of what I was working in anyway.
• 5:32
  SPEAKER_03: So uh, you know, does it feel like to people that it just raised the performance bar, but it it fooled us?
• 5:41
  SPEAKER_02: Well, there's a lot in that question.
• 5:43
  SPEAKER_02: Let's start with what the promise was.
• 5:47
  SPEAKER_02: And I think probably people rem more or less remember this, because of course we're it's not that long ago, right?
• 5:54
  SPEAKER_02: Really, for most people, the AI world began on November 30th, 2022, with the launch of Chat GPT 3.5.
• 6:06
  SPEAKER_02: And that was for you know people that don't recall this, that was the thing where it kind of like came out of the lab and was like, hey, look at us.
• 6:14
  SPEAKER_04: Yeah.
• 6:15
  SPEAKER_02: And it it felt in that moment as though we had there was before that date and there was after that date.
• 6:26
  SPEAKER_02: Rather like, now I'm really gonna age myself here, but rather like when Google first appeared, Google the search engine first appeared, and until that we'd been using things like uh well, Yahoo's search engine, Alta Vista, Alt Al Alta Vista, which itself was, you know, seemed seemed pretty impressive and various Yeah, Ask G's and crawlers and all that kind of stuff.
• 6:52
  SPEAKER_02: Anyway, so yeah, so when Google came out and you could type something in, and you seemed to appear to be able to type more or less anything in, and you got a response back in you know, half a second or something.
• 7:03
  SPEAKER_02: Yeah, that it was like, oh, everything has changed.
• 7:07
  SPEAKER_01: Right.
• 7:08
  SPEAKER_02: When Chat GPT came out, and you just seemed to be able to open the thing up and ask it stuff, and it seemed to like be really smart and answer questions, and remember everything that had gone before it, there have been these sort of attempts, right?
• 7:23
  SPEAKER_02: There was the you know, the piece of crap that was uh that was Siri and the piece of crap that was that that was Alexa.
• 7:31
  SPEAKER_03: Alexa, yeah.
• 7:31
  SPEAKER_02: And very and and obviously those things are you know are continuing to uh continue to evolve, but they were so wrong so often, and all of a sudden there was this thing that appeared to be pretty right quite often.
• 7:44
  SPEAKER_04: Right.
• 7:45
  SPEAKER_02: That felt like, regardless of any external marketing or anything they did, that felt like, oh, now suddenly it's like Star Trek.
• 7:58
  SPEAKER_02: But then, you know, rapidly they're raising money.
• 8:02
  SPEAKER_02: There's companies like Microsoft and others behind them, there are marketing teams behind them, and now we're being painted a picture, and the picture we're being painted is really type a couple of things in or speak to it like Scotty does the computer in in Star Trek.
• 8:19
  SPEAKER_02: Yeah, sit on a beach and do your work in a fraction of the time than you could have done previously.
• 8:25
  SPEAKER_02: And it was almost like for people that have read four-hour work week and things like that, it sort of felt like, okay, now we're in the world, we're finally in the world where everybody gets to live Tim Ferris' dream.
• 8:37
  SPEAKER_02: Yeah, and we're all we're all gonna be working four-hour work weeks, and we're gonna be monstrously productive, and all of our stress is gonna go away, and we're gonna be able to spend more time with our family, and our work is gonna be better as well, right?
• 8:49
  SPEAKER_02: And as I say, we did some of that to ourselves, right?
• 8:52
  SPEAKER_02: We did some of it through breathless social media and all of this kind of stuff as uh um as well, but we created that impression in ourselves that this would that this was the new world and there was massive upside to this new world.
• 9:06
  SPEAKER_02: Now, because we're human beings, right, there are gonna be some human beings who look at that and go, yeah, I can sit on a beach.
• 9:13
  SPEAKER_02: There are gonna be some human beings who are like, oh man, you know, I used to be able to do X in eight hours, and now I can be doing 10x in eight hours.
• 9:22
  SPEAKER_01: Yeah.
• 9:23
  SPEAKER_02: And then there are gonna be still more that are gonna look at that and be frightened and we're gonna want to hide into a corner.
• 9:31
  SPEAKER_02: Now, if you're an employee that is an individual contributor and work isn't your be all and end all, then you're probably very attracted to the idea of getting everything done in two hours and uh sitting on a beach for the rest.
• 9:45
  SPEAKER_02: If you're a manager inside that same organization, you may well be looking at that and going, ooh, now even the most unpro even the most unproductive person in my company can do massively more than they would have been able to previously.
• 10:02
  SPEAKER_02: And so there was this, we saw very rapidly, there have been multiple articles written about this, uh, we can link to a BBC one in this.
• 10:10
  SPEAKER_02: We saw very rapidly that it accelerated the difference of expectations between how employees were looking at work and how their managers and their companies were looking at work.
• 10:21
  SPEAKER_02: And both of them were kind of unrealistic.
• 10:24
  SPEAKER_02: Um and and both of them were based on the fact that neither party even really understood what this technology was and how it was transforming work.
• 10:30
  SPEAKER_02: So you've got the employee thinking I can work for two hours, you've got the employer thinking you can do ten times as much, and there's not a lot of meaning in the middle there.
• 10:37
  SPEAKER_02: And as a result, you wind up where we are, kind of.
• 10:40
  SPEAKER_03: And the incentives are completely misaligned.
• 10:42
  SPEAKER_02: Yeah.
• 10:43
  SPEAKER_03: Like polar opposites almost.
• 10:46
  SPEAKER_02: Polar opposites.
• 10:47
  SPEAKER_03: Yeah.
• 10:47
  SPEAKER_03: Right.
• 10:47
  SPEAKER_02: And this isn't the first time that has happened, but we there is a lot of evidence out there to show that employee and employer expectations around work have never been further apart.
• 10:58
  SPEAKER_02: Um, and this is a very strong contributing factor to that.
• 11:01
  SPEAKER_03: That's fascinating.
• 11:03
  SPEAKER_03: Um let's move to thinking about um what we're expecting from these things, optimization, etc.
• 11:17
  SPEAKER_03: Are we optimizing individual tasks but quietly breaking the human operating system required to perform them?
• 11:26
  SPEAKER_02: I like that as a um I like that as a concept.
• 11:31
  SPEAKER_02: Yeah, I mean, we are certainly dramatically changing what it takes to be really, really good at um at work, right?
• 11:40
  SPEAKER_02: A different set of skills need uh needed in order to be really good at work.
• 11:45
  SPEAKER_02: Um and of course that's what I'm trying to, a large part of what I'm trying to document in the AI Ready Human is what are those capabilities and how do you go how do you go about building those capabilities and highlighting the fact that most of us don't uh you know don't work on them on a day-to-day basis.
• 12:01
  SPEAKER_02: But yeah, your phrasing of that is really is really interesting.
• 12:05
  SPEAKER_02: You've brought in this you've brought in a new technology, but as you've done it, you brought in a new set of expectations, you've brought in a new uh a new sort of how, what, when, where, and why of work.
• 12:20
  SPEAKER_02: All very rapidly together.
• 12:23
  SPEAKER_02: You've done so without training people properly, managers or employees.
• 12:32
  SPEAKER_02: And hopefully it's clear to our audience by now when I say training, I do not mean prompt engineering, right?
• 12:37
  SPEAKER_02: I mean like you've done so without training people how to thrive, uh, how to thrive in this uh in this new world.
• 12:45
  SPEAKER_02: And so in that sense, right, the if you think of the if you take your operating system analogy, right?
• 12:55
  SPEAKER_02: If you think of your operating system being the how work gets done operating system that sort of sits in a uh in a company and you have to, and you as a human have to interact with it, that's just been changed.
• 13:08
  SPEAKER_04: Yeah.
• 13:09
  SPEAKER_02: Right.
• 13:09
  SPEAKER_02: And so you don't know how to operate effectively in in that environment.
• 13:18
  SPEAKER_02: And at the same time, your manager is reading, I don't know, you know, 10 prompts that'll make you work more efficiently in their feed, and they're like, oh, this is so easy.
• 13:32
  SPEAKER_02: I don't know why everybody doesn't get it.
• 13:34
  SPEAKER_02: And one of the most interesting things about writing the book is that both in researching the book and in feedback from people that have read the book, I've I've heard from managers who were thinking, I thought this was easy, and just my people were not getting it.
• 13:58
  SPEAKER_02: And they've shifted from that to now I realize it isn't easy, and I also realize that I wasn't getting it.
• 14:07
  SPEAKER_02: And so now they're looking at it, they're re-evaluating how they work with it, but then also reevaluating how their employees work with it as well as a as a result of all this.
• 14:19
  SPEAKER_02: And that can be the starting point for getting closer alignment between the manager and the employee of what we're actually dealing with.
• 14:28
  SPEAKER_03: Do you think that the the the burnout rate and the burnout sense is also related to um this kind of battle that's happening with people psychologically, where like if you're being asked to um use AI and you're also hearing that AI is going to replace you, you know, it's kind of like we've created the the robot equivalent of the person who that you hire and you know, like the young kid, you're like they're they're here to replace me, I know it.
• 15:05
  SPEAKER_03: And now I gotta train them, like that resentment that would build, you know, like yeah, and and I so we gotta be looking at it that way too.
• 15:13
  SPEAKER_02: Like Oh my goodness, yeah.
• 15:15
  SPEAKER_03: Yeah.
• 15:15
  SPEAKER_02: Well, let's take a let's take a step back and think about the think about the factors that are that are leading into this.
• 15:21
  SPEAKER_02: I should mention, by the way, we haven't covered it really, but but for anybody that's listening and going, actually, is is this even real?
• 15:27
  SPEAKER_02: Is this actually happening?
• 15:28
  SPEAKER_02: Are people are are people more burned out as a result of AI?
• 15:32
  SPEAKER_02: I've written an article about this, we'll put it in the show notes.
• 15:35
  SPEAKER_02: But the but the there's a definitive answer to that is yes.
• 15:38
  SPEAKER_02: Not everybody, but there are but but absolutely there is the case that you know AI-driven burnout is a real thing, it is happening, and it's happening inside every every organization.
• 15:48
  SPEAKER_02: So if you're in an organization right now, it's happening inside your organization.
• 15:52
  SPEAKER_02: Um and there are many contributing factors that are that are associated with this.
• 16:00
  SPEAKER_02: Um one of them is simply that the the amount of work that we're doing on a day-to-day basis is more, and humans do have a limit, right?
• 16:11
  SPEAKER_02: Um, and doing lots and lots of things faster um can be more tiring, right?
• 16:19
  SPEAKER_02: Um people burned out, you know, the the classic I Love Lucy sketch, you know, on the factory line, right?
• 16:25
  SPEAKER_04: Yeah.
• 16:25
  SPEAKER_02: People burned out when they were having to do lots and lots of things in a period of time and those things sort of felt this felt samey.
• 16:33
  SPEAKER_02: Well, if you're ironically, if you're interacting with AI, then a lot of the stuff you are doing is kind of samey.
• 16:39
  SPEAKER_02: You see a problem, you go to a prompt, you evaluate what's in the prompt, you having having evaluated what's in the prompt, you think of the next prompt, you ask the next prompt, and you're almost like a factory worker interacting uh back and forth with AI.
• 16:51
  SPEAKER_02: But but the but yes, you're getting lots done.
• 16:54
  SPEAKER_02: Getting lots done isn't always excessively it isn't excessively motivating in its own right.
• 16:59
  SPEAKER_02: The second thing is that, and this was fascinating that we I saw in my own research, is that one of the ways in which you can theoretically be most productive with AI is to find all the things, go across all the processes that you you do on a day-to-day basis, and all the things that you do, and then extract all of those things where AI is as good as you or nearly as good as you, and give all that stuff to AI.
• 17:29
  SPEAKER_02: Now, in the process, and I saw this definitively over several months with the people that I was studying, um, many of those people take that to such an extreme that now AI is doing all of the things that they actually quite enjoyed.
• 17:43
  SPEAKER_02: And so it's changed the flip of the ratio of stuff that you do at work that you like and stuff that you do at work that you find mundane, and it's increased the amount of stuff that you do that's mundane and got rid of the amount of stuff that you that um that you like.
• 17:59
  SPEAKER_02: And then another, a third really important factor is what you just alluded to there, which is it's the AI equivalent of um the beatings will continue until morale improves.
• 18:11
  SPEAKER_02: It is the it's it it is you have literally got people standing up on stage, leaders of leaders of AI companies proudly standing up on stage and saying, yeah, you know, companies are gonna need one quarter of the workforce that they've needed previously, or one fifth of the workforce they've needed previously.
• 18:30
  SPEAKER_02: And people are watching that, right?
• 18:32
  SPEAKER_02: And then you've got um leaders talking incessantly about okay, we're gonna get to 5x or 6x or 10x what we were uh what we were doing uh previously, and they see the attrition that that's there right now.
• 18:53
  SPEAKER_02: Even the companies that are not firing people, some of those people companies are like, okay, this, you know, Julie is retired, we won't replace Julie.
• 19:03
  SPEAKER_02: They see that, and they're not stupid, right?
• 19:07
  SPEAKER_02: They look at that and go, okay, well, this is the first wave.
• 19:10
  SPEAKER_02: What's the next wave?
• 19:11
  SPEAKER_02: What's the next wave?
• 19:12
  SPEAKER_02: And so you are, if if you're an employee in that environment and you're not getting clear, rational communication from your leadership as to what the strategy is and what the role of the humans in the organization is associated with the strategy, you're gonna make up your own stories.
• 19:33
  SPEAKER_02: And so when you do that, right, if you're a 35-year-old, I don't know, individual contributor or middle manager who was doing pretty well and understood the human relationships inside your company and kind of had an idea as to where you're gonna be in three years' time or five years' time, and now you're looking forward in five years going, Well, I don't know what the world of work is gonna look like, but I don't, and nor do I know whether I have a role in it.
• 20:00
  SPEAKER_02: And I could be looking forward five years and I work super hard every day, and yet I might be sat on my backside watching TV with no with no ability to be able to do anything meaningful in three or four or five years' time.
• 20:13
  SPEAKER_02: That's very dispiriting.
• 20:14
  SPEAKER_04: Yeah.
• 20:14
  SPEAKER_02: And so lack of like like lack of motivation, particularly lack of being able to see a good future for yourself, yeah, is closely related to Bennett.
• 20:24
  SPEAKER_03: Yeah, that's why I think your book is so great because it's it's it it's the answer to dealing with that in a lot of ways.
• 20:34
  SPEAKER_03: Like it because the other choice is ignore it, like put blinders on and hope you last it doesn't come for you.
• 20:41
  SPEAKER_03: Like it it's it's really interesting.
• 20:44
  SPEAKER_03: It's I think that's something that that you know really is is great about the way you set up your book.
• 20:49
  SPEAKER_03: Um so you talked uh in the article about two different ways that people have been using AI, and I'd like to hear a bit more on that, and the time savers versus the quality of work uh aspect.
• 21:03
  SPEAKER_03: Uh I thought that was pretty fascinating because um the combination of the two changes the dynamic as well.
• 21:12
  SPEAKER_02: So yeah, well, you know me well enough to know that I've had a uh I have an off-on relationship with the concept of mindset, right?
• 21:20
  SPEAKER_04: Yeah.
• 21:21
  SPEAKER_02: And so um just because uh I'm not diminishing mindset in in any way or grow you know, the power of uh of a of a growth mindset or anything along those lines, but what I have seen um in multiple organizations is the misuse of things like growth mindset and um and the weaponization of things like uh like growth mindset.
• 21:46
  SPEAKER_02: Um but none of that diminishes the fact that the mindset that you bring to work as an individual, um, how you choose to think about your work.
• 22:00
  SPEAKER_02: Does make a big difference in the results that you drive, the motivation to continue working and all of that kind of stuff.
• 22:10
  SPEAKER_02: And also that as humans, um, with a few exceptions, as humans, by and large, we have a we have more control over that than we than we think we do.
• 22:26
  SPEAKER_02: And this is this is very interesting territory because we know that about you know 90 in at least in the 90s percent of what we do is driven by sort of underlying brain functions, and then we have this executive function that justifies what we already did.
• 22:38
  SPEAKER_04: Yeah.
• 22:39
  SPEAKER_02: But it doesn't mean that the executive function, in other words, how we choose to think about things, can't have an influence in the in the other direction.
• 22:46
  SPEAKER_02: And so I was very interested in that, and I was very interested when I was researching the book as to was there correlation between how people the mindset that people brought to AI and their use of AI and what their ultimate relationship was with both AI work, I guess, or three things, right?
• 23:14
  SPEAKER_02: AI itself, work, and their broader life.
• 23:18
  SPEAKER_04: Yeah.
• 23:19
  SPEAKER_02: And correlation and causation, you know, who knows?
• 23:26
  SPEAKER_02: Who knows exactly what uh um where the chicken and the egg is here, but it is true to say that um people by and large when they start working with AI, they think of it in one of two different ways.
• 23:45
  SPEAKER_02: They think of it as primarily as a time saver, namely, sweet, I can get this much done in much less time, or I've got this amount of time and I can get more done.
• 24:02
  SPEAKER_02: It's the productivity view of the world.
• 24:05
  SPEAKER_02: And then there's this second category of people that are sweet, I could get a B, but using AI, I can get an A plus.
• 24:19
  SPEAKER_02: And by the way, there are gonna be people who are listening to this going, yeah, I do both, I do both.
• 24:25
  SPEAKER_02: And that's true, right?
• 24:26
  SPEAKER_02: But there but but there is a primary motivator for different kinds of people.
• 24:31
  SPEAKER_02: And so when I started researching the book, I it was just really kind of based on instinct as much as anything else, but I was but I was thinking, I wonder if those people will have different outcomes.
• 24:43
  SPEAKER_02: Right.
• 24:44
  SPEAKER_02: And sure enough, when when I went to the same groups of people like um uh several months later, the by and large, the quality improvers, the people that said, Oh, I could get a B and now I can get an A plus, those people were had a continued healthy relationship with with AI and were enjoying their relationship with AI and were doing better work and doing better as individuals.
• 25:12
  SPEAKER_02: And the people that were like the the time savers associated with AI, they got themselves on a treadmill that was going faster and faster and faster.
• 25:22
  SPEAKER_02: They were on that I love Lucy factory floor, yeah, right, and they were not.
• 25:27
  SPEAKER_02: They were their relationship with AI had soured typically, they didn't really know how to get out of it.
• 25:33
  SPEAKER_02: Um and of course, if you think about it, if you're the quality improver, the quality improver also extends to the processes and the way you use AI.
• 25:42
  SPEAKER_02: It's very inquisitive in nature.
• 25:43
  SPEAKER_02: You're like, oh, what if I used AI in the middle of a process versus the start of a process?
• 25:48
  SPEAKER_02: What would that do?
• 25:49
  SPEAKER_02: How would that change things?
• 25:51
  SPEAKER_02: And so you start up so rather counterintuitively, the quality improvers were also mostly quite good at getting time savings as well, because that was a quality of process that they were focused on.
• 26:03
  SPEAKER_02: The time savers were simply driving themselves, their managers, their spouses crazy with their inability to be able to do anything other than turn up the speed of the of the treadmill.
• 26:17
  SPEAKER_03: Yeah, and you said something in there, I had a pull, I pulled the quote out.
• 26:20
  SPEAKER_03: Uh it's it's as if they're I think it was related to this, papering over their weaknesses temporarily.
• 26:27
  SPEAKER_03: Like yeah.
• 26:28
  SPEAKER_02: Yeah, that's it's certainly it's it's certainly absolutely related to that.
• 26:33
  SPEAKER_02: And as we get into one of the other uh um later on with one of the other questions, we'll talk about um whether whether generative AI is actually any good at anything.
• 26:45
  SPEAKER_02: And it will um and you'll start to see how that connects to uh to this idea of sort of using AI to paper to paper over cracks.
• 26:58
  SPEAKER_02: But it is it's interesting, right?
• 27:00
  SPEAKER_02: Because if you are a if you truly are a quality improver, then you're gonna be naturally cautious of things like, okay, I don't know anything about X, so I'll just ask AI and assume it's true.
• 27:13
  SPEAKER_02: Right.
• 27:14
  SPEAKER_04: Right, right.
• 27:14
  SPEAKER_02: Those those things are things that the quality improver wouldn't necessarily do.
• 27:21
  SPEAKER_02: The quality improver, for example, when AI, when they're using AI to help them research, right, and AI says, oh, you know, um, here's what we know about this topic, they may ask AI for sources, AI might provide them sources, and then they would even go the extra step and click on those links and see are those links valid, valid links, are they, if let's say they're looking at studies, are those studies real studies?
• 27:47
  SPEAKER_02: They might even read through the studies, right?
• 27:50
  SPEAKER_02: And so they would they go down the rabbit hole in a very productive way until they get to the point where they've actually gained real understanding, excuse me, understanding versus um just simply getting it out quick and moving on to the next thing.
• 28:04
  SPEAKER_03: Yeah.
• 28:05
  SPEAKER_03: Yeah, it's it's it is a very interesting time to watch this happening.
• 28:10
  SPEAKER_03: Like I I I don't think uh I would have predicted necessarily the the natural extension to where we are now, even being around for the dawn of the internet.
• 28:20
  SPEAKER_03: Yeah.
• 28:21
  SPEAKER_02: You know, like well and one thing I would sorry, go ahead.
• 28:23
  SPEAKER_03: Uh it it's just it's fascinating to watch this uh the speed that this is happening.
• 28:31
  SPEAKER_03: It's it's well yeah, exactly.
• 28:34
  SPEAKER_02: And one thing that I I think we still don't know the answer to, Matt, is we've we've lived in a world many of the managers who listen to this podcast and the leaders who listen to the podcast might be screaming at the screaming at us right now.
• 28:52
  SPEAKER_02: Because we've lived in a world for the longest time whereby managers and leaders have been kind of annoyed at their employees Um who are the quality improvers, right?
• 29:06
  SPEAKER_02: Because they're like, no, no, no, I what I need here is good enough, right?
• 29:10
  SPEAKER_02: I don't need I don't need perfection.
• 29:13
  SPEAKER_02: And what you're telling me, Paul, is that the quality the quality improvers are the ones that can sustainably use this stuff versus the uh versus the time savers.
• 29:28
  SPEAKER_02: And yes, I mean I'm just simply saying that's what that that's that's what is the connection is to burnout here.
• 29:35
  SPEAKER_04: Yeah, yeah.
• 29:36
  SPEAKER_02: But the thing is that I think this is gonna lead us to a it to a very interesting ultimate spot where the reason that this whole idea of you know good enough was good enough was that we had many pieces of work where the cost from getting from good enough to really good was way too high.
• 30:02
  SPEAKER_02: And so this is one of the areas where AI potentially changes that, right?
• 30:08
  SPEAKER_02: AI potentially changes the fact that getting from good to excellent is the cost of that is much lower, and ultimately what differentiates your business are the things that you're excellent at, not the things that you're good enough at.
• 30:28
  SPEAKER_02: And so we could potentially we could use AI to gain excellence at low cost in areas that we were unable to gain excellence before, and that might change what our view of this sort of like good enough world is.
• 30:47
  SPEAKER_02: So those are some of the this is where it does, where the question itself does get incredibly interesting and complicated.
• 30:54
  SPEAKER_03: Yeah.
• 30:55
  SPEAKER_03: I I this might be a little bit of a tangent, but you brought you brought something up, and I I wanted to mention this to you, because it's kind of in in this realm, if we're talking about people burning out now and training their, you know, training AIs with their their knowledge.
• 31:14
  SPEAKER_03: Um this is sort of the Library of Alexandria of knowledge that we're building, and we're gonna put a lot into it if we don't have people coming up behind that also know how it got to the Library of Alexandria.
• 31:32
  SPEAKER_03: Like, like we're putting a lot of faith in we're gonna have the people who know the experts are gonna train and check the AI, and once we finish that task, now we don't have to worry about it anymore.
• 31:44
  SPEAKER_03: And there's not gonna be the junior people who have to take 20 years to learn something, right?
• 31:49
  SPEAKER_03: So they're never they're never gonna get a chance to burn out because they're never even gonna be never even gonna be in there to begin with.
• 32:00
  SPEAKER_03: And I wonder if that gaining of experience part helps us avoid burnout because we're learning the whole time and we're now cutting out even that aspect that we were kind of enjoying.
• 32:21
  SPEAKER_03: Does that resonate?
• 32:22
  SPEAKER_02: It does.
• 32:23
  SPEAKER_02: Um it's funny because burnout is such a complicated It's such a complicated thing.
• 32:32
  SPEAKER_02: Um if you are learning and growing without fear, you're not gonna, you're generally not gonna burn out.
• 32:41
  SPEAKER_02: Right.
• 32:42
  SPEAKER_02: Right?
• 32:42
  SPEAKER_02: So you're learning people like most people, they like to learn and grow, right?
• 32:48
  SPEAKER_02: They like to have they like the feeling of being better today than they were yesterday and being better tomorrow than they were today.
• 32:56
  SPEAKER_02: Um and you're exactly right.
• 32:59
  SPEAKER_02: I mean, if we it's it's sort of like we there are there were two forms of burnout that that have historically existed.
• 33:11
  SPEAKER_02: And the one is that you're just the factory worker on the factory line and you're burned out because everything is so boring, and and yet, you know, if you make a mistake and you accidentally, you know, you're so bored and you're so tired that you accidentally put your hand in the machine at the wrong time is going to rip your arm off.
• 33:34
  SPEAKER_02: And so there's that sort of boredom associated with fear that was kind of like the the uh one of the original sort of forms of burnout.
• 33:43
  SPEAKER_02: And then today's form of burnout is is that you're constantly trying to operate in a world of deep ambiguity and uncertainty.
• 33:57
  SPEAKER_02: You don't know what good looks like, you don't know if you're working, you don't know if you're working hard enough, you don't know if, even if you're working as hard as you feel you can, you don't know if you're gonna have a job, you don't know any of those things.
• 34:09
  SPEAKER_02: So fear is always a contributing factor, no matter whether it was fear combined with being mundane or fear associ uh fear and uncertainty, whether those things were combined.
• 34:21
  SPEAKER_02: They either way they kind of they kind of lead to burnout.
• 34:23
  SPEAKER_04: Yeah.
• 34:24
  SPEAKER_02: But you're exactly right.
• 34:27
  SPEAKER_02: If we're feeding this machine, right, I'll call it the Bodleian Library because I'm from Oxford and that and they have every published book in the world, but it's a um, but it's a it's the idea that we're just gonna plug everything into the machine and then the machine is gonna do everything, and I don't know, we congratulate ourselves because we were the ants that that did all the work to make that happen.
• 34:53
  SPEAKER_02: That's not that's not fun.
• 34:55
  SPEAKER_02: So yeah, I think that I think the things are uh I think they are kind of like deeply, deeply associated with each other.
• 35:03
  SPEAKER_02: And so it's a and then to your point, right?
• 35:08
  SPEAKER_02: It is putting an awful lot of faith, right?
• 35:10
  SPEAKER_02: This thing that we're whatever it is that we're building, and who knows what it is that we're you know, everything that we're building inside companies, yeah, the body of knowledge that is being built inside uh inside companies, the the data that we have in the internal data pools that we have, that AI is gonna magically interrogate all this wonderful unstructured data that's gonna mean that you know we don't have to rely on you know on our retirees to understand this this stuff.
• 35:45
  SPEAKER_02: We we do not know.
• 35:47
  SPEAKER_02: We do not know yet um how reliable that's gonna be.
• 35:53
  SPEAKER_03: It's not gonna be because you've got to think it's not.
• 35:56
  SPEAKER_03: Because w we're not gonna be there to check it.
• 36:00
  SPEAKER_03: And the people that come after us will assume we were right, otherwise, how did it get into the machine?
• 36:05
  SPEAKER_03: It's almost like this is uh as we're talking, I'm like, oh man, this is this is wild.
• 36:12
  SPEAKER_02: Well, and I also wonder about like there are lots of things that we haven't fully considered.
• 36:17
  SPEAKER_02: Historical bias, for example.
• 36:19
  SPEAKER_04: Yeah.
• 36:19
  SPEAKER_02: Like one of the we took we talked in the episode where we were going through the through the book, and we talked about the fact that that AI doesn't have context in the same way that humans have context.
• 36:32
  SPEAKER_02: Right?
• 36:32
  SPEAKER_02: AI, the the fact that that we're having a you and I are having a meandering discussion that is in a different spot now than it was a little while ago, right?
• 36:43
  SPEAKER_02: How do we train AI?
• 36:44
  SPEAKER_02: We train AI on historical data, right?
• 36:48
  SPEAKER_02: And how how can AI fully know, in the way that humans know, that the behaviors that we exhibited a year ago, two years ago, three years ago, are not the behaviors that are gonna get us where we need to be.
• 37:09
  SPEAKER_01: Yeah.
• 37:09
  SPEAKER_01: Right?
• 37:10
  SPEAKER_01: Yeah.
• 37:10
  SPEAKER_02: And because by definition, you make these things smart based on based on historical data.
• 37:18
  SPEAKER_02: Yeah.
• 37:18
  SPEAKER_02: And then you do wind up, this is particularly challenged in certain industries, you're in the life sciences industry, for example, you do wind up having to think very carefully and very seriously about bias in the in the original data sets and what that leads you to do.
• 37:37
  SPEAKER_02: And are you using AI in such a way that by definition you're amplifying the bias that is that is in those data sets.
• 37:45
  SPEAKER_02: And so it's that phrase about the model, all models are wrong, but some models are more useful than others.
• 37:52
  SPEAKER_02: Yeah.
• 37:53
  SPEAKER_02: Right?
• 37:54
  SPEAKER_02: Uh, that's my I've got two favorite uh phrases about models.
• 37:58
  SPEAKER_02: That's one.
• 37:59
  SPEAKER_02: The second phrase about models is the best model of a cat is a cat.
• 38:03
  SPEAKER_02: And what and what both of those are getting at is really the same the same thing, right?
• 38:10
  SPEAKER_02: Which is that there's a real world where real things happen, and then there is a instantiation of that, a digital instantiation of that, which is wrong, and it's just a question of how wrong.
• 38:26
  SPEAKER_03: How wrong, yeah.
• 38:28
  SPEAKER_02: And and there's a point at which it being wrong by a bit is okay, and then there's a point beyond that where it's not okay.
• 38:39
  SPEAKER_04: Yeah.
• 38:40
  SPEAKER_02: And if there are little to no humans in the mix to determine what is the case, which of those two things is the case, then we have a problem.
• 38:51
  SPEAKER_02: Now that's gonna, that's related to the topic we're gonna talk about next uh next time, which is AI making us dumber.
• 38:59
  SPEAKER_04: Yeah.
• 38:59
  SPEAKER_02: But it's a but it's re it's relevant here also, right?
• 39:03
  SPEAKER_02: And it is this undercurrent that that we've been discussing here, this undercurrent that we're developing a we're developing a world where amazing things can happen, right?
• 39:17
  SPEAKER_02: We'll be able to do things, we already can do things that are that have been unimaginable before unimaginable before.
• 39:27
  SPEAKER_02: And if we do it in the right way, we're creating a world where we're combining AI and humans in exactly the right way to achieve order of magnitude better results.
• 39:39
  SPEAKER_04: Yeah.
• 39:40
  SPEAKER_02: Right?
• 39:40
  SPEAKER_02: So the AI systems that are, for example, working in the background, expert systems to help human beings leverage and improve their own expertise to achieve significantly greater results.
• 39:54
  SPEAKER_02: Those those things are incredible.
• 39:56
  SPEAKER_02: And they're gonna and they have the potential to uh to lift us up.
• 40:00
  SPEAKER_02: But they have to do it if humans are gonna humans can provide value in this picture, and if humans are going to provide value in this picture, then they can't be burned out.
• 40:09
  SPEAKER_02: And so that's the that's the that's the recipe that we have to make here, which brings us back to the idea of like how do leaders behave to make sure that the humans are not burned out.
• 40:20
  SPEAKER_02: Yeah, how do leaders hire, how do the how leaders, how do leaders retain, who do you hire, how do you hire, how do you train, how do you how do you manage, how do you communicate, how do you educate.
• 40:33
  SPEAKER_02: All of these things are vital if we're gonna have highly engaged human beings who are not burned out, who are gonna contribute to this future that everybody wants.
• 40:43
  SPEAKER_03: And and they're exposed in that um how they define these things can have a huge impact on how, like, and I'm saying management is exposed by defining the the goalposts.
• 40:59
  SPEAKER_03: Yeah.
• 41:01
  SPEAKER_03: They contribute directly to the expectation.
• 41:03
  SPEAKER_03: Like if it's one thing to say, all right, we we have AI now, so you know, FTE footprint now can be cut in half.
• 41:13
  SPEAKER_03: Or what is this thing?
• 41:16
  SPEAKER_03: Like, do you approach it, how they approach it is going to directly impact everybody's burnout rate.
• 41:22
  SPEAKER_02: Oh my goodness, yeah.
• 41:24
  SPEAKER_03: Yeah.
• 41:24
  SPEAKER_02: And the stakes are hugely high too.
• 41:27
  SPEAKER_02: Because one thing that we do know, and we've known for really since at least for the last 40, 50 years, is that there's something it's kind of it's it's kind of strange when you think about it, right?
• 41:40
  SPEAKER_02: But there's something different about a IT system or computer making a mistake and a human making a mistake.
• 41:50
  SPEAKER_02: And particularly when it makes an expensive mistake.
• 41:54
  SPEAKER_02: So um think about a world where technology has been very, very closely involved in um in systems that like keep us safe, like the increasing prevalence of self-driving cars, for example.
• 42:12
  SPEAKER_02: In the round, in general, um when once we get to a world, particularly once we get to a world where all cars are self-driving, then that's a much more predictable world than where you've got some self-driving cars on the road, you've got some people, and you've got a bunch of people on the road, some of whom are tired, some of whom are drunk, some of whom are drugged, um, and that they're all trying to play by the same rules and somehow and somehow work things out.
• 42:41
  SPEAKER_02: The that world in which we live today, right, there's lots of car crashes every year.
• 42:46
  SPEAKER_02: And quite a few people die in those car crashes.
• 42:50
  SPEAKER_02: But when an autonomous vehicle kills people, that's an entirely different, that's an entirely different world, right?
• 43:02
  SPEAKER_02: And the company most likely is gonna be held accountable in some way, society is gonna reject it, you know, etc.
• 43:11
  SPEAKER_02: etc.
• 43:12
  SPEAKER_02: I mean, for everything that you think about Tesla and so on, it's not killing 20,000 self- Tesla self-driving cars are not killing 20,000 people a year.
• 43:20
  SPEAKER_02: If they were, we would shut it down in an instant.
• 43:25
  SPEAKER_02: So now take that to sort of the general corporate world.
• 43:28
  SPEAKER_02: Let's say that you're a senior leader and you decide, and you're a senior leader of a company that has safety systems or systems that can like really impact people's lives.
• 43:41
  SPEAKER_02: Uh that could be their financial security, it could be their health, it could be their education or something like that.
• 43:46
  SPEAKER_02: But let's say something that's life threatening, right?
• 43:48
  SPEAKER_02: So, in other words, where something goes wrong in your company, people die.
• 43:51
  SPEAKER_04: Yeah.
• 43:51
  SPEAKER_02: Right?
• 43:52
  SPEAKER_02: If you if you pull out 40% of your staff, and then a year later, After you've pulled out 40% of your staff, a inappropriately governed or inappropriately transparent AI system, in other words, an AI system that has a lack of transparency, there's a black box, as we have to call it, makes a mistake and three humans die.
• 44:28
  SPEAKER_02: What do you think is going to happen to the stock price of your company?
• 44:33
  SPEAKER_02: Now a human could make that mistake and it would be and it would be entirely different.
• 44:38
  SPEAKER_02: But because you you went through the step of consciously pulling the humans out of the mix, consciously relying on AI and saying we're going to do that because it's going to yield better results, et cetera, et cetera, and then that fails, and then people die, you're in a world of hurt.
• 44:55
  SPEAKER_02: So everybody in the organization, whether choosing to think about it or not, are dealing with things that are quite risky, and which leads us all the way back to where we're at with burnout, right?
• 45:07
  SPEAKER_02: Burnout is typically higher when there is immediate risk or risk in the background.
• 45:15
  SPEAKER_02: And so there is back, there is background risk associated with these decisions that that people are making.
• 45:21
  SPEAKER_02: That could be a background risk of, oh, I fired a bunch of people and then I made a mistake.
• 45:26
  SPEAKER_02: You know, as we speak here today, uh Jack Dorsey and his uh in his company is has uh decided they're getting rid of about 50% of their stuff.
• 45:34
  SPEAKER_02: That's a that's a can we operate leanly with with 50% less people.
• 45:39
  SPEAKER_02: But then there are the ones where it's like, no, we fired, we hired, we fired people and then people are are suffering.
• 45:46
  SPEAKER_02: Customers are suffering, um, or you know, people that we serve are suffering in some way, shape, or form.
• 45:53
  SPEAKER_02: And then that it that background risk is something that leads to burnout.
• 45:58
  SPEAKER_03: Yeah.
• 45:58
  SPEAKER_03: It strikes me that the the responsibility uh of success has to come both directions.
• 46:08
  SPEAKER_03: Yeah.
• 46:08
  SPEAKER_03: Like leadership is is on the hook to be educated and understand and have the right expectations, and the the teams have to see the big picture, see the value, and and try to it's an incentives um thing that that you choose you have to choose the right incentive.
• 46:29
  SPEAKER_02: And and if the incentive lights on here, Matt.
• 46:33
  SPEAKER_02: We're getting into darkness.
• 46:34
  SPEAKER_02: All right.
• 46:35
  SPEAKER_02: It's very cool.
• 46:36
  SPEAKER_02: All right.
• 46:36
  SPEAKER_03: If all the uh incentives aren't in line uh or they're the wrong ones, it could really drive burnout.
• 46:43
  SPEAKER_03: So yeah.
• 46:44
  SPEAKER_03: Um let's let's bring it uh down to sort of uh a point here.
• 46:50
  SPEAKER_03: If if AI is changing how we work every minute of the day, and and we can just accept that we're all forced to think about it at this point if we're if we're in that world.
• 47:02
  SPEAKER_03: First, what is it doing to how we actually think?
• 47:07
  SPEAKER_03: Yeah.
• 47:07
  SPEAKER_03: And uh what should we do knowing what we know coming out of this uh this first part of the big question?
• 47:20
  SPEAKER_03: What is it what is what is the answer to the big question, is my question.
• 47:27
  SPEAKER_02: So I'll we've we've we've gone around this so much, so I'll draw us back to what the question actually was in the first place.
• 47:34
  SPEAKER_02: So why is AI making uh some of us more burned out?
• 47:38
  SPEAKER_02: And I love the way you're closing it down here, because you're like, okay, well, this burnout is clearly related to how we think.
• 47:45
  SPEAKER_02: Um we've talked about the fact that you know the mindset you bring seems to have an impact on this.
• 47:52
  SPEAKER_02: And then the broader question is what is it doing in terms of how it makes us think?
• 47:57
  SPEAKER_02: I think there's a couple of dimensions to it, one of which we'll cover next week, which is the, you know, is it making us um is it making us dumber?
• 48:04
  SPEAKER_02: But related to the burnout question specifically, um I think what it's doing in terms of how it makes us think, is that if we're using it in certain ways, and if the incentives are in our company to use it in those ways, namely we can we can all do 10 times what we did previously.
• 48:30
  SPEAKER_02: So everybody go and do, remember that phrase, do more with less, right?
• 48:35
  SPEAKER_02: This is our we asked you forever to do more with less.
• 48:39
  SPEAKER_02: This is our tool to allow you to do more with less.
• 48:43
  SPEAKER_02: If it's framed in that way, if it's framed in that way in the minds of employees, in managers, and so on and so forth, um that in itself is a recipe for burnout.
• 48:59
  SPEAKER_02: And so what I would encourage people to do is flip the script on this entirely.
• 49:08
  SPEAKER_02: Um what we want to get to is a point where you have highly empowered employees who are using AI to do the best, most rewarding work that they've ever done.
• 49:30
  SPEAKER_02: And I don't know how many in a in a company, you don't necessarily know how many of those people you need, but what you do know is you need them, right?
• 49:40
  SPEAKER_02: Because those people, those people that are highly empowered, once they're highly empowered and they're doing great work in this new world, by definition they're adaptable because the new world is different to the is different to the old world.
• 49:58
  SPEAKER_02: And so what you've got then is like an army of highly adaptable people that then can adapt to wherever the technology, uh, wherever the technology is going.
• 50:08
  SPEAKER_02: And so it comes all the way back to the to the main theme of the book, I suppose, which is that in order to be successful in this new world, you must be highly adaptable.
• 50:19
  SPEAKER_02: And coming with that is the idea of having the autonomy to do great work, mastering this new world of work, mastering it, and then doing it for a reason, right?
• 50:32
  SPEAKER_02: Doing it for a positive reason, giving people a positive reason to do the work that they're gonna do.
• 50:38
  SPEAKER_02: And so I think that you start with that idea that we're gonna create, we're gonna create these people.
• 50:48
  SPEAKER_02: We're gonna create a workforce of these people.
• 50:50
  SPEAKER_02: We're gonna hire people that think about AI and work with AI in that way.
• 50:56
  SPEAKER_02: We're gonna train people to work with AI in that way.
• 51:02
  SPEAKER_02: And then that's gonna be our collection of people that are gonna take our organization to new heights and in the process take themselves to new heights as well.
• 51:15
  SPEAKER_02: And that's the antithesis of burnout.
• 51:18
  SPEAKER_02: So it's a flipping, it really is sort of a flipping of the script from everybody do more with less to let's create amazingly adapted people that will that because they're amazingly adaptable, because they have purpose to what they are doing, they are doing more.
• 51:36
  SPEAKER_02: So you don't start with do more, you start with create the people, and then those people themselves do more.
• 51:42
  SPEAKER_02: By definition, those people are not burned out.
• 51:44
  SPEAKER_03: They'll know what to do with the AI instead of they will know what to do with the AI.
• 51:48
  SPEAKER_03: Yeah.
• 51:49
  SPEAKER_03: Right.
• 51:49
  SPEAKER_03: Yeah.
• 51:50
  SPEAKER_02: That's a good place to wrap it.
• 51:51
  SPEAKER_02: That yeah, exactly.
• 51:52
  SPEAKER_02: It changes your it it changes how you how you think about it.
• 51:56
  SPEAKER_02: It raises certain things like uh like, for example, employee development, it raises that up much more versus some of the other things.
• 52:02
  SPEAKER_03: Yeah.
• 52:03
  SPEAKER_03: Well, this has been awesome.
• 52:04
  SPEAKER_03: I had a great, great time talking about this.
• 52:06
  SPEAKER_03: I can't wait for the never for the next 7BQ AI.
• 52:11
  SPEAKER_03: Uh and uh we'll talk to you soon.
• 52:14
  SPEAKER_03: Thanks, Bob.
• 52:15
  SPEAKER_00: Yeah, thank you.
• 52:18
  SPEAKER_00: Humanity Working is produced by Matt Neal.
• 52:21
  SPEAKER_00: If you liked this episode, please take a moment to rate and review us on your favorite podcast platform.
• 52:27
  SPEAKER_00: It makes a huge difference in helping others find us.
• 52:30
  SPEAKER_00: And don't forget, you can find all episodes and get our weekly newsletter at humanityworking.net.
• 52:38
  SPEAKER_00: See you next time.