Latest / The SOM Podcast / Future of Work, AI, and Occupational Health: Navigating Change and Uncertainty

Transcript

Future of Work, AI, and Occupational Health: Navigating Change and Uncertainty · The SOM Podcast

• 0:00
  Speaker 1: This podcast is sponsored by orchid Live. Find out more
• 0:04
  Speaker 1: at orchidlive dot com.
• 0:18
  Speaker 2: Hello, and I'll come to the societyat of Ocupational Medicine Podcasts.
• 0:21
  Speaker 2: I'm dot SORIS Center GP and Occupational Health physician and
• 0:24
  Speaker 2: I'm here with Cat Bernard.
• 0:26
  Speaker 3: So yes, hello everybody. My name is Cat Barnard, and
• 0:29
  Speaker 3: I am a director at a company called Working the Future,
• 0:34
  Speaker 3: which is a future of work advisory firm. In short,
• 0:38
  Speaker 3: we track and analyze a range of trends and we
• 0:41
  Speaker 3: help businesses understand those trends and prepare for and adapt
• 0:47
  Speaker 3: to very different work futures.
• 0:49
  Speaker 2: And that's a great place to start because we're going
• 0:52
  Speaker 2: to be talking about artificial intelligence work and health, and
• 0:57
  Speaker 2: it's something that I'm very interested in for our specialty
• 1:00
  Speaker 2: because I think as well as changing what and how
• 1:04
  Speaker 2: our specialty will do and function, AI is obviously going
• 1:08
  Speaker 2: to change workplaces across the board and therefore change the
• 1:12
  Speaker 2: bread and butter of what we see in our clinics.
• 1:15
  Speaker 2: So I just love to get your kind of high
• 1:18
  Speaker 2: level take on that to start with, and the kind
• 1:21
  Speaker 2: of trends that you're seeing are made.
• 1:23
  Speaker 3: One of the things that I just want to point
• 1:25
  Speaker 3: out is that much of the desk BACE research that
• 1:29
  Speaker 3: we track and analyze pertains most specifically to people who
• 1:34
  Speaker 3: work in offices. Although that said, I think, you know,
• 1:38
  Speaker 3: it's quite an interesting thought experiment, isn't it to imagine
• 1:42
  Speaker 3: the professions or the roles that people undertake as part
• 1:47
  Speaker 3: of their work lives that aren't going to be impacted
• 1:50
  Speaker 3: by artificial intelligence in some way, shape or form. Because
• 1:54
  Speaker 3: obviously lots of people don't work in offices. If you
• 1:56
  Speaker 3: think about people that work in healthcare profession, if you
• 1:58
  Speaker 3: think about people that work in the educational profession, and
• 2:01
  Speaker 3: yet the issue of mainstream generative AI is a key
• 2:05
  Speaker 3: issue for those people as well. So I tend to
• 2:08
  Speaker 3: kind of try and broaden the aperture and think about
• 2:12
  Speaker 3: where are the professions where AI will not feature? And
• 2:18
  Speaker 3: really loosely, I can think about things like, you know,
• 2:20
  Speaker 3: hairdressing or construction work, or let's say agriculture or horticulture,
• 2:28
  Speaker 3: or you know, professions where you use your hands. That said,
• 2:32
  Speaker 3: how will the administrative side of those professions get changed?
• 2:35
  Speaker 3: But Lucy speaking, when we're talking about the impacts of
• 2:39
  Speaker 3: generative AI on people at work, we tend to talk
• 2:43
  Speaker 3: about people who are working in offices who are increasingly
• 2:48
  Speaker 3: expected to use digital tools as productivity resources to help
• 2:54
  Speaker 3: them do their work quicker, better, faster.
• 2:56
  Speaker 2: So there's kind of like access of exposure, and that
• 3:00
  Speaker 2: depends a little bit on industry. It depends a little
• 3:02
  Speaker 2: bit on occupation within industry. It probably also depends a
• 3:06
  Speaker 2: little bit on whether you're a junior or experienced higher
• 3:09
  Speaker 2: and also whether AI at this point, because of course
• 3:13
  Speaker 2: the technology continues to evolve, whether it's automating tasks or
• 3:18
  Speaker 2: augmenting human impact. And I think you've kind of illustrated
• 3:21
  Speaker 2: that really well with the examples you've given across the economy.
• 3:25
  Speaker 2: So we'll be thinking, I guess a little bit more
• 3:27
  Speaker 2: about those desk based roles in this forum.
• 3:31
  Speaker 3: Yeah, but equally, I guess, you know, sort of kicking
• 3:33
  Speaker 3: it back into your area of expertise and specialism. One
• 3:37
  Speaker 3: of the challenges that we've got with the whole advent
• 3:40
  Speaker 3: of AI is that it's a one size fits all
• 3:44
  Speaker 3: kind of label where actually there's a whole raft of
• 3:48
  Speaker 3: different subcategories underneath it. So if I think about, for instance,
• 3:52
  Speaker 3: and I am absolutely the least qualified person in the
• 3:56
  Speaker 3: room here, that if I think about the way in
• 3:58
  Speaker 3: which artificial intelligence can be used in let's say the
• 4:03
  Speaker 3: field of radiography to analyze X rays and other images
• 4:09
  Speaker 3: allegedly far more accurately than the human mind can. That
• 4:13
  Speaker 3: is a very different application of AI to his co
• 4:18
  Speaker 3: pilots in this building that occupies a thousand workers, go
• 4:22
  Speaker 3: and use it.
• 4:23
  Speaker 2: I agree. So there's different use cases within the economy too,
• 4:27
  Speaker 2: And I think that kind of brings me to this
• 4:30
  Speaker 2: question around what are you seeing on the ground, Because
• 4:35
  Speaker 2: what we see from our practice and also in the literature,
• 4:39
  Speaker 2: and I think the literature is catching up here. It's
• 4:42
  Speaker 2: not like there's a huge amount of evidence generation on
• 4:45
  Speaker 2: what is actually happening to the workforce or the humans
• 4:50
  Speaker 2: who are deploying AI more in their work. We're seeing
• 4:54
  Speaker 2: that it sometimes isn't that smooth going, and there's a
• 4:58
  Speaker 2: change element that's challenging. There are these other risk factors
• 5:02
  Speaker 2: for work relata stress that also can kind of come
• 5:05
  Speaker 2: out down the line when AI is thrown into the mix.
• 5:08
  Speaker 2: For example, people without managerial experience suddenly supervising agents, or
• 5:13
  Speaker 2: people beginning to use AI to unlock skills that they
• 5:17
  Speaker 2: didn't previously have but then requiring human supervision or people
• 5:21
  Speaker 2: finding that they can be less boundaried about work because
• 5:24
  Speaker 2: task initiation is so easy when you're speaking to a chatbot.
• 5:28
  Speaker 2: So I just would love to hear your take on
• 5:31
  Speaker 2: what you're seeing with the organizations you work with.
• 5:34
  Speaker 3: Yeah, okay, So what might be useful is to try
• 5:37
  Speaker 3: and apply some kind of a historical lens if we can.
• 5:43
  Speaker 3: If you think about you know, we're in twenty twenty six,
• 5:46
  Speaker 3: the mainstream Internet was born. If you were birthed in
• 5:53
  Speaker 3: the nineteen nineties. The naughties was the decade where we
• 5:58
  Speaker 3: saw the arrival of social media and the advent of
• 6:02
  Speaker 3: cloud computing, which meant that organizations could outsource their technology
• 6:08
  Speaker 3: environments to the cloud. We know that the arrival of
• 6:15
  Speaker 3: cloud computing and the concept of big data, the idea
• 6:22
  Speaker 3: that every organization owns vast volumes of proprietary information held
• 6:28
  Speaker 3: either in the cloud or previously on its own servers.
• 6:32
  Speaker 3: We know that digital transformation has been a big thing
• 6:36
  Speaker 3: in business for probably the last ten years. There's some
• 6:41
  Speaker 3: really interesting data points out there about digital transformation. Most notably, McKinsey,
• 6:48
  Speaker 3: the big management consultant, has recorded pretty consistently that seventy
• 6:53
  Speaker 3: percent of transformation programs fail And originally, when that data
• 6:58
  Speaker 3: first came out, it was like full stop, they fail.
• 7:01
  Speaker 3: But actually when you dig around in the background of that,
• 7:05
  Speaker 3: you start to see an emergent picture, which is that
• 7:08
  Speaker 3: they fail typically because of the human lack of adoption.
• 7:14
  Speaker 3: And typically, in my experience, failure to adopt happens because
• 7:19
  Speaker 3: there is an incoherent narrative. There aren't sensible rules of
• 7:26
  Speaker 3: engagement that are presented. A story isn't told about why
• 7:31
  Speaker 3: that adoption needs to happen, and what is going to
• 7:35
  Speaker 3: happen to the humans that work within the organization, how
• 7:39
  Speaker 3: will they be impacted? And the one thing that I
• 7:41
  Speaker 3: think is true about AI adoption, which is a massive
• 7:46
  Speaker 3: cause of concern but also a massive opportunity for those
• 7:50
  Speaker 3: working in occupational health, is that there isn't a coherent
• 7:54
  Speaker 3: story around what is going to happen to the jobs
• 7:57
  Speaker 3: that people do, and if people are anxious that their
• 8:01
  Speaker 3: jobs are unsafe because of the arrival of generative AI
• 8:06
  Speaker 3: in an economic backdrop where we have just experienced cost
• 8:11
  Speaker 3: of living crisis after cost of living crisis since twenty twenty,
• 8:16
  Speaker 3: with things like you know, the start of the Ukraine War.
• 8:19
  Speaker 3: Now what's going on in the Middle East supply chain
• 8:22
  Speaker 3: shortages that happened as a result of COVID. You know,
• 8:25
  Speaker 3: we seem to be living through a time where the
• 8:28
  Speaker 3: global economy is becoming increasingly unsable and uncertain, which is
• 8:34
  Speaker 3: causing a lot of people to be very, very worried
• 8:37
  Speaker 3: about their financial futures, and there is data out there
• 8:40
  Speaker 3: to support it. You know that financial wellbeing is one
• 8:44
  Speaker 3: of the pillars of well being overall. Well If you
• 8:47
  Speaker 3: add into the mix a lack of coherent narrative around
• 8:51
  Speaker 3: job security arising from the adoption or the expected adoption
• 8:57
  Speaker 3: the required adoption of generative AI, it stands to reason
• 9:01
  Speaker 3: that you know, people are less likely to adopt tools
• 9:07
  Speaker 3: if there is a belief that those tools could render
• 9:10
  Speaker 3: them redundant.
• 9:12
  Speaker 2: Makes total sense, and I can see the anxiety element
• 9:15
  Speaker 2: there and how that may impact people's ability to get
• 9:19
  Speaker 2: on board with change and impact their productivity and impact
• 9:22
  Speaker 2: their health and the relationship between their health and their work.
• 9:25
  Speaker 2: I think part of the challenge, obviously, is the technology
• 9:28
  Speaker 2: is changing all the time, and it must be so
• 9:31
  Speaker 2: difficult for organizations to think ahead about how they will
• 9:38
  Speaker 2: communicate with their workforce and how this will look and
• 9:42
  Speaker 2: I would also say that before we even get to replacement,
• 9:46
  Speaker 2: the process is still challenging. The kind of shift in
• 9:50
  Speaker 2: what people do day to day and their shift in
• 9:53
  Speaker 2: thinking and their shift in workplace tasks seems to come
• 9:58
  Speaker 2: with some kind of you pacial health fallout. I'm really
• 10:02
  Speaker 2: interested in what you see with organizations between whether you
• 10:07
  Speaker 2: see and I've seen a bit of this few and
• 10:09
  Speaker 2: far between, but kind of systematic adoption of technology with
• 10:14
  Speaker 2: really clear ground rules for the workforce and access to
• 10:19
  Speaker 2: tools and guidelines, and really encouraging building out time within
• 10:24
  Speaker 2: the week for people to experiment with AI, whether they're
• 10:27
  Speaker 2: a technical hire or not, versus kind of giving people
• 10:31
  Speaker 2: access and encouraging them to use AI versus no real
• 10:35
  Speaker 2: narrative around it and more of an ad hoc approach
• 10:38
  Speaker 2: where some interested people might be bringing in some tools
• 10:42
  Speaker 2: and other parts of the workforce aren't at AOL. And
• 10:45
  Speaker 2: whether you see that spectrum in your practice and how
• 10:49
  Speaker 2: you find each of those in supporting the workforce.
• 10:53
  Speaker 3: Yeah, I mean every point that you raised there is
• 10:56
  Speaker 3: a really significant one. The data is kind of it's
• 11:01
  Speaker 3: warm data, like it comes out in drips and drabs,
• 11:03
  Speaker 3: and you know you'll see a piece of research undertaken
• 11:06
  Speaker 3: over here and a piece of research undertaken over there.
• 11:09
  Speaker 3: And of course research is quite sexy in the twenty twenties.
• 11:13
  Speaker 3: So one of the things that we have to be
• 11:14
  Speaker 3: discerning about is what is credible research? Right? Is it
• 11:18
  Speaker 3: research that has been pulled together for marketing purposes or
• 11:22
  Speaker 3: is it bona fidey, well grounded, well structured research. That's
• 11:26
  Speaker 3: point number one. And then I think it is clear
• 11:31
  Speaker 3: from what we have seen that the level of investment
• 11:37
  Speaker 3: into the digital tools is not accompanied or complemented by
• 11:43
  Speaker 3: an equivalent amount of investment into upskilling the human beings
• 11:48
  Speaker 3: who are then expected to use those tools. That's one point.
• 11:53
  Speaker 3: Another point, and there's some lovely research that's been undertaken
• 11:57
  Speaker 3: by Megan Rights and her research partner John Higgins. So
• 12:01
  Speaker 3: Meghan is a professor at I think Oxford these days,
• 12:06
  Speaker 3: and they've been exploring the concept of spaciousness and how
• 12:11
  Speaker 3: little time is granted to people at work to think
• 12:15
  Speaker 3: and to play and to experiment. And as you pointed out,
• 12:19
  Speaker 3: you know, playing is actually a primary way in which
• 12:23
  Speaker 3: we learn as social you know creatures, right, So there
• 12:27
  Speaker 3: isn't the time that is being allocated to just explore
• 12:32
  Speaker 3: and noodle around and figure out there certainly doesn't seem
• 12:36
  Speaker 3: to be any coherent accepted standards of kind of minimum
• 12:44
  Speaker 3: viable usage around you know, ethics and privacy and confidentiality
• 12:50
  Speaker 3: and data sovereignty and fundamentally big questions that do have
• 12:56
  Speaker 3: a long term impact for the healthy functioning of organizations, communities, societies, countries, etc.
• 13:04
  Speaker 3: So all of those things kind of weave in and
• 13:07
  Speaker 3: out of one another. In the absence of a credible
• 13:12
  Speaker 3: story or a strategic narrative or a framework for sensible adoption,
• 13:18
  Speaker 3: people make up stories for themselves, and without clear parameters,
• 13:24
  Speaker 3: they'll bring their own tech to work, they'll play in
• 13:27
  Speaker 3: ways that perhaps are suboptimal for the organization and the
• 13:31
  Speaker 3: objectives of the organization. So I'm really curious about the
• 13:36
  Speaker 3: role that communication plays as we navigate into the future
• 13:43
  Speaker 3: and what strikes me. And there was some data that
• 13:45
  Speaker 3: we picked up in one of the more recent research
• 13:48
  Speaker 3: updates that we produced for the Institute of Internal Communication.
• 13:52
  Speaker 3: Seventy seven percent of UK workers are pretending to know
• 13:57
  Speaker 3: more about generative AI than they actually do. This is
• 13:59
  Speaker 3: a according to some data that was published by a
• 14:02
  Speaker 3: company called plural Site, and in that same roundup. We
• 14:06
  Speaker 3: also discovered that ninety one percent of senior executives are
• 14:10
  Speaker 3: exaggerating their knowledge of AI whilst demanding that their colleagues
• 14:16
  Speaker 3: down the organization adopt it. So it just feels like
• 14:20
  Speaker 3: it's just a confusing, mucky murky mess until we start
• 14:25
  Speaker 3: being clear about how we use the tech and what
• 14:30
  Speaker 3: our expectations are. And also we allow sandboxed experiments for
• 14:36
  Speaker 3: people to experiment and make mistakes, because that's how we learn.
• 14:40
  Speaker 3: If you think about how we learn as human beings,
• 14:43
  Speaker 3: we learn through making mistakes, don't we. So we have
• 14:46
  Speaker 3: to provide them the environments for people to sit with
• 14:50
  Speaker 3: the technology and learn in a structured way, but we
• 14:55
  Speaker 3: don't seem to be providing any of that situational you know,
• 15:02
  Speaker 3: the environment for people to discover for themselves how to
• 15:06
  Speaker 3: use the tech. And the final thing that I want
• 15:09
  Speaker 3: to say on this, Lara, because I think this is
• 15:11
  Speaker 3: a really big point that isn't being addressed. If you
• 15:15
  Speaker 3: think about the way that most people have been hired
• 15:20
  Speaker 3: into their jobs, mostly what happens the process starts with
• 15:26
  Speaker 3: the creation of a job description and then the budgetary
• 15:30
  Speaker 3: sign off, Yes, we have the money to bring on
• 15:33
  Speaker 3: somebody to do this work. Somebody somewhere has written a
• 15:37
  Speaker 3: set of roles and responsibilities, and the roles and responsibilities
• 15:40
  Speaker 3: are then accompanied by a set of desired skills and
• 15:44
  Speaker 3: then sometimes accompanied by a set of nice to have skills,
• 15:48
  Speaker 3: but that is typically the makeup of a job description.
• 15:52
  Speaker 3: You then decide, well, that looks like an interesting job.
• 15:55
  Speaker 3: I'm going to go and apply for it. You attend,
• 15:57
  Speaker 3: you know, you send your CV, your CV gets shortlisted.
• 16:01
  Speaker 3: You a tend a range of interviews, you have a
• 16:03
  Speaker 3: range of conversations with your future employer, and you know, finally,
• 16:07
  Speaker 3: happy days, you are offered the job. To what extent
• 16:11
  Speaker 3: have you constructed that job? You personally haven't constructed that job.
• 16:16
  Speaker 3: You've applied for a job that you know that you
• 16:19
  Speaker 3: have got the aptitude, the skills, etc. To deliver on,
• 16:22
  Speaker 3: but you haven't actually created the job. We've now got
• 16:26
  Speaker 3: generative AI coming into organizations, and what needs to happen
• 16:32
  Speaker 3: is that we collectively need to determine which tasks can
• 16:37
  Speaker 3: be handed off to an AI. But most people don't
• 16:42
  Speaker 3: know which task can be handed off because they've never
• 16:45
  Speaker 3: constructed their own jobs. It's quite a thing, isn't it.
• 16:48
  Speaker 3: You can do the work. That's one thing. You can
• 16:51
  Speaker 3: do the work. You know what your roles and responsibilities
• 16:53
  Speaker 3: are on a day and day out basis. But actually
• 16:56
  Speaker 3: taking a step back and a helicopter view of why
• 17:00
  Speaker 3: do I do the things that I do? Why is
• 17:02
  Speaker 3: my work stacked up in this way? What constitutes a
• 17:05
  Speaker 3: work stack? What constitutes a workflow? That's a completely different
• 17:10
  Speaker 3: level of thinking that most people have had zero experience of.
• 17:16
  Speaker 3: So to land to drop a technology into a workplace
• 17:20
  Speaker 3: and expect people to be able to reverse engineer which
• 17:24
  Speaker 3: elements of their work can be handed off to an
• 17:28
  Speaker 3: AI that also needs training and it needs sensible parameters
• 17:33
  Speaker 3: for safe adoption, none of us have got that expertise, right,
• 17:38
  Speaker 3: very very few of us have got this expertise. So
• 17:41
  Speaker 3: to expect the technology to land in a way that
• 17:46
  Speaker 3: brings about the efficiency is absurd.
• 17:50
  Speaker 2: No, I agree with you. A role, of course, is
• 17:52
  Speaker 2: a very big part of the work related stress picture.
• 17:56
  Speaker 2: Role clarity. When am I doing what's not my role? Hey? Changed?
• 18:00
  Speaker 2: I thought I was meant to be doing this. We
• 18:03
  Speaker 2: know that that causes work related stress. But it's a
• 18:06
  Speaker 2: very interesting lens that you bring there around. Okay, well,
• 18:10
  Speaker 2: that's going to happen, But also who's meant to be
• 18:13
  Speaker 2: determining these roles? Can it just evolve organically or not?
• 18:19
  Speaker 2: And I suppose again, I'm just thinking about the challenge
• 18:22
  Speaker 2: of uncertainty with this technology and which humans find hard.
• 18:27
  Speaker 2: Humans find uncertity really hard, and what that means for
• 18:30
  Speaker 2: organizations who are trying to leverage the technology and communicate
• 18:36
  Speaker 2: that uncertainty and still deploy it. Do you have any
• 18:40
  Speaker 2: thoughts on that, and like how the uncertainty could be managed?
• 18:44
  Speaker 3: So I think the sooner the organizations start addressing the
• 18:48
  Speaker 3: human factors the psychology of change, the better. I don't
• 18:53
  Speaker 3: think we can expect healthy, robust, sustainable AI adoption and
• 19:00
  Speaker 3: we start to address the human factors. And I defer
• 19:03
  Speaker 3: to you, I am not a clinician, right but having
• 19:07
  Speaker 3: worked with high performing teams for several decades, I've got
• 19:11
  Speaker 3: an elevated interest in psychology and how people perform well
• 19:16
  Speaker 3: at work, on what happens when they don't, and what
• 19:19
  Speaker 3: the reasons are for why they don't. One of the
• 19:21
  Speaker 3: things that strikes me is that one of the most
• 19:24
  Speaker 3: useful models of behavior change that I have come across
• 19:29
  Speaker 3: is the procesca declemente trans theoretical model of change, which
• 19:34
  Speaker 3: was devised in the nineteen seventies, I think, and it
• 19:39
  Speaker 3: seems to be the case that it's been used with
• 19:41
  Speaker 3: great efficacy in behavior change programs including smoking cessation, weight loss,
• 19:48
  Speaker 3: lifestyle management, addiction recovery, et cetera. Etc. Maybe I was
• 19:54
  Speaker 3: thinking about us recording this conversation, right, but I've just
• 19:57
  Speaker 3: in the last week. I've just gone round I don't
• 19:59
  Speaker 3: know dozen people that I know who work directly in
• 20:03
  Speaker 3: business transformation or change communication to inquire whether or not
• 20:10
  Speaker 3: that model of change features in any transformation programs, and honestly,
• 20:16
  Speaker 3: I was met with blank stares right across the board.
• 20:19
  Speaker 3: Nobody had heard of that model. And so it strikes
• 20:22
  Speaker 3: me as quite peculiar that from an organizational point of view,
• 20:27
  Speaker 3: businesses approach change as if it was a modular process
• 20:33
  Speaker 3: to be managed rather than a deeply personal psychological process.
• 20:41
  Speaker 3: Do you know the guy Andrew Huberman. He has a
• 20:44
  Speaker 3: podcast called Huberman Lab. He's a neuroscientist, I think, based
• 20:48
  Speaker 3: in the States, and one of his findings was that
• 20:51
  Speaker 3: it takes typically between eighteen and two hundred and fifty
• 20:56
  Speaker 3: four days for a human to successfully change habit. Well,
• 21:01
  Speaker 3: I don't think that's factored into any change programs either.
• 21:04
  Speaker 2: So exactly, And it's the pace of change and the
• 21:08
  Speaker 2: kind of dynamics of the change. It's not like you
• 21:12
  Speaker 2: could set out on a journey around AI transformation and
• 21:16
  Speaker 2: six months later that journey would look the same, because
• 21:19
  Speaker 2: of course the technology is evolving all the time, so
• 21:22
  Speaker 2: I think there's a real challenge in communicating and managing
• 21:26
  Speaker 2: those variables.
• 21:28
  Speaker 3: But that kind of.
• 21:28
  Speaker 2: Six changes of six stages of change model that you
• 21:32
  Speaker 2: outline will probably be familiar to many of the listeners
• 21:36
  Speaker 2: to this podcast, because, as you say, it's something that
• 21:40
  Speaker 2: we come across in medical training to manage addictions. And
• 21:43
  Speaker 2: I think it's a very interesting concept around how we
• 21:46
  Speaker 2: could possibly use that to support the workforce and support
• 21:52
  Speaker 2: the organization. And I'm thinking again about what you said
• 21:56
  Speaker 2: about research and again how organizations can conduct research internally
• 22:01
  Speaker 2: to help feedback on what's going on on the ground
• 22:05
  Speaker 2: with the humans as they adjust to working with more AI.
• 22:08
  Speaker 2: So I think there's plenty to be done here, and
• 22:12
  Speaker 2: lots of occupational health clinicians and teams can weigh in on.
• 22:16
  Speaker 2: But what would your kind of takeaway be for an
• 22:20
  Speaker 2: occupational health team who are seeing a little bit of
• 22:24
  Speaker 2: fallout on human AI interaction in their workstreams. Where do
• 22:30
  Speaker 2: you think we should pitch our efforts.
• 22:32
  Speaker 3: So counter to what seems like a very dominant Silicon
• 22:39
  Speaker 3: Valley skewed narrative about AI adoption, I believe that successful
• 22:47
  Speaker 3: AI adoption is contingent on the human capacity for that adoption.
• 22:54
  Speaker 3: And you're absolutely right. It's not singularly a once and
• 22:59
  Speaker 3: done option thing, because by the time you've adopted generative
• 23:04
  Speaker 3: AI slavor A, you know, flavors B, C, and D
• 23:08
  Speaker 3: will have arrived. It's also the case that AI adoption
• 23:13
  Speaker 3: is being floated at the same time the organizations are
• 23:18
  Speaker 3: having to think about carbon emission reduction or reducing their
• 23:22
  Speaker 3: dependency on fossil fuels because of what's going on in
• 23:25
  Speaker 3: the Middle East now, or you know, shifting attitudes towards work,
• 23:29
  Speaker 3: what are going to be the skills of the future. Like,
• 23:32
  Speaker 3: there's just so many flavors of change floating around in
• 23:36
  Speaker 3: the ether all at once. It's not singularly AI adoption
• 23:40
  Speaker 3: that needs to be overseen. And the other thing going
• 23:44
  Speaker 3: back to that change cycle is we know because the
• 23:48
  Speaker 3: cycle sets it out, it's iterative right, and there's always
• 23:52
  Speaker 3: the possibility of relapse. So we need to be thinking
• 23:56
  Speaker 3: instead of, I think, instead of assuming that change at
• 24:01
  Speaker 3: work is a linear, step by step process by which
• 24:06
  Speaker 3: we will arrive at a set outcome, I think, sooner
• 24:10
  Speaker 3: we can start thinking about the process of change as
• 24:13
  Speaker 3: being a cyclica process that is deeply, deeply human. Like
• 24:21
  Speaker 3: we know because the data is showing us that programs
• 24:25
  Speaker 3: of change fail because of a lack of human uptake.
• 24:29
  Speaker 3: So how might we as practitioners reframe transformation as being
• 24:36
  Speaker 3: a very human process that is cyclica rather than linear.
• 24:43
  Speaker 3: And how might we use what we know about the
• 24:47
  Speaker 3: psychology of change to encourage and invite conversations in organizational
• 24:53
  Speaker 3: settings that allow people to understand the basis by which
• 24:59
  Speaker 3: humans change habits and become a bit more comfortable with
• 25:02
  Speaker 3: both the process of habit change. Like once you can
• 25:06
  Speaker 3: see it, you can you tend to identify it within yourself.
• 25:10
  Speaker 3: It tends to ring true once somebody has outlined a
• 25:14
  Speaker 3: process to you. So there's that piece and then there's
• 25:17
  Speaker 3: the piece around. It's okay to fail. Failures are part
• 25:21
  Speaker 3: of learning. It's also okay to not feel okay about
• 25:25
  Speaker 3: the process of change, like we are emotional creatures, aren't we,
• 25:29
  Speaker 3: And some days will feel up and other days will
• 25:32
  Speaker 3: feel down. And I think perhaps if we were a
• 25:34
  Speaker 3: tiny bit more compassionate with ourselves and with our colleagues,
• 25:38
  Speaker 3: we might be able to make more progress rather than
• 25:41
  Speaker 3: this very mechanical way of thinking about organizational change, which
• 25:45
  Speaker 3: is somehow the human is an automaton that will just
• 25:48
  Speaker 3: do these things and then these other things will happen
• 25:51
  Speaker 3: as a result.
• 25:52
  Speaker 2: I really like that kind of analogy of moving away
• 25:55
  Speaker 2: from linear towards a more cyclical model of what's going on.
• 25:59
  Speaker 2: And I think for people working in occupational health there's
• 26:01
  Speaker 2: a kind of almost synergy there with like an audit
• 26:04
  Speaker 2: cycle or things they can do to sort of support organizations,
• 26:09
  Speaker 2: gather work and health data, and be a part of
• 26:13
  Speaker 2: that iterative process in a more evidence based way. So
• 26:16
  Speaker 2: I think that's a great takeaway. What's an interesting conversation.
• 26:19
  Speaker 2: Thank you so much for coming on and for Chastis.
• 26:22
  Speaker 3: Thank you very much for inviting me. It's pleasure.
• 26:33
  Speaker 1: This podcast is sponsored by orchid Live. Orchid Live is
• 26:37
  Speaker 1: a specialist occupational health software used by a number of
• 26:41
  Speaker 1: some members to run every aspect of their occupational health operations.
• 26:46
  Speaker 1: They hold healthcare records for over one million UK workers
• 26:49
  Speaker 1: and work with both in house OH teams and OH providers.
• 26:54
  Speaker 1: You can find out more at orchidlive dot com