Latest / In the Loop / What does "human in the loop" really mean?

Transcript

What does "human in the loop" really mean? · In the Loop

• 0:04
  Abbie Day: Hello everyone, and welcome to this week's episode of In the Loop, where we bring you in-depth discussions about AI in the insurance industry.
• 0:11
  Abbie Day: I'm Abby Day, Senior Reporter for Insurance Insider, and today we will be diving deeper into the podcast namesake to learn a bit more about the phrase so often heard when discussing AI, human in the loop.
• 0:22
  Abbie Day: Joining me is Instanda CEO Tim Harcastle, and we will be discussing what the phrase human in the loop really means, the difference between human-led and human-enabled AI, and what approach is best for your business.
• 0:34
  Abbie Day: As well as whether the insurance industry will ever use AI without a human in the loop.
• 0:38
  Abbie Day: Without further ado, let's get started.
• 0:41
  Abbie Day: So this week we are deep diving into human interaction with AI with a very exciting guest.
• 0:53
  Abbie Day: We have Tim Harcastle, CEO and co-founder of Instanda.
• 0:57
  Abbie Day: Hello and welcome, Tim.
• 0:58
  Tim Hardcastle: Great to see you.
• 0:59
  Abbie Day: So we are looking to understand more about what it means to have a human in the loop, kind of going off of our podcast namesake.
• 1:06
  Abbie Day: So maybe we can start for any listeners who may not know Instanda or know who you are with a quick introduction to the company.
• 1:12
  Tim Hardcastle: Okay, Instanda was created just over 10 years ago.
• 1:16
  Tim Hardcastle: We're part of the Insure Tech Wave, and it's the world's few, if not the only, no code policy admin platform.
• 1:23
  Tim Hardcastle: We serve MGAs, tier one, tier two, three carriers in the US, uh, UK and Europe, and Australia.
• 1:31
  Tim Hardcastle: We currently have around about 100 clients using the platform, and obviously we're very privileged to work with them because what we do is exceptionally important to their business success.
• 1:40
  Tim Hardcastle: And we're at this great inflection point, which we'll talk about around how the future of technology will evolve in the world of AI.
• 1:47
  Abbie Day: Amazing.
• 1:47
  Abbie Day: So we've had conversations about AI in the past, and one of the conversations we've had that's really stuck out to me is the meaning of the words human in the loop.
• 1:55
  Abbie Day: We've spoken about human-enabled, human-led AI.
• 1:59
  Abbie Day: Maybe we could start with the phrase human in the loop, and you could explain to me exactly what it really means.
• 2:05
  Abbie Day: It's a phrase we hear thrown around so much.
• 2:08
  Tim Hardcastle: Well, like so many um parts of the AI landscape that everyone is experiencing interacting with, I think things are evolving and moving quite quickly.
• 2:15
  Tim Hardcastle: From when we were talking, we talked about human in the loop.
• 2:18
  Tim Hardcastle: And I think I'd look at it through three different ways to express the human interaction with AI.
• 2:23
  Tim Hardcastle: So simplistically, first one would probably what I would describe as human-led, where the person does the work and decides, and AI prepares and suggests.
• 2:36
  Tim Hardcastle: And I think many people would recognize that's how the casual user might interact with AI today.
• 2:42
  Tim Hardcastle: Human in the loop, I would characterize that as where the AI does the work and the person checks or approves it at set points.
• 2:50
  Tim Hardcastle: And then the third way that you can look at the human and AI interaction is where human-enabled AI works on its own within hard limits, so well-defined limits.
• 3:02
  Tim Hardcastle: So people set the rules, people watch the results and then deal with the exceptions.
• 3:08
  Tim Hardcastle: So there's three layers the human-led, the human in the loop, and what I'd call human-enabled.
• 3:12
  Abbie Day: Okay, so maybe we could go a bit more into how to decide which way you want your company to go.
• 3:18
  Abbie Day: So we can start maybe with the underwriting side.
• 3:21
  Abbie Day: Is it fair to kind of assign one type to one type of company?
• 3:26
  Tim Hardcastle: Yeah, I think each level is trading control against scale.
• 3:31
  Tim Hardcastle: So human-led, where the human is making the decision, the AI is providing the draft, if you like, and a bit of a suggestion.
• 3:38
  Tim Hardcastle: It gives the best judgment.
• 3:40
  Tim Hardcastle: So for an underwriting perspective, it would be something where you're dealing with complex, high consequence risks, but it's the slowest part of, let's say, the decision-making process to scale for obvious reasons.
• 3:54
  Tim Hardcastle: In contrast, human-enabled is where, as I said, the AI is doing the work, the human is looking at the exceptions.
• 4:02
  Tim Hardcastle: So, for example, you get real scale with this, for example, pulling out data out of documents.
• 4:07
  Tim Hardcastle: Obviously, you need to have good solid foundations around clean data, got your set parameters, and you need to record everything that AI is doing.
• 4:14
  Tim Hardcastle: But essentially, AI is doing the work for you, and you're, as I said, you're looking at the exceptions.
• 4:20
  Tim Hardcastle: So if you contrast those two things, human-led, where the human is making the decision and AI is providing some of the drafting and suggestions, versus human-enabled, you can see the difference between the control and the scale.
• 4:31
  Tim Hardcastle: And then the one in the middle, human in the loop, you get throughput with some control because the AI may be triaging submissions, drafting a quote, for example, inside an underwriter's authority, and then the underwriter approved.
• 4:44
  Tim Hardcastle: So that the human in the loop in this context is the one in the middle from a scale and control perspective.
• 4:49
  Tim Hardcastle: And I think an organization should look at the kind of activities, tasks, and decisions it needs to make.
• 4:56
  Tim Hardcastle: And it would decide whether it's going to be human-led, as he as in very heavily oriented around the underwriter making these complex decisions or human-enabled, where the AI is doing a lot more of the processing and the humans are looking at just the exceptions.
• 5:08
  Abbie Day: Okay, so maybe we could talk about the teams themselves.
• 5:11
  Abbie Day: Does the team you're working with, the team you're managing, change the way you would approach this as well?
• 5:17
  Tim Hardcastle: It's a really interesting point because I think when we start to reconsider and reframe, if you like, the workplace environment, the structure almost of what the future will hold and what some companies are experiencing even now, what we can see is jobs, in the way I've just described those interactions between the AI and humans, jobs are shifting from doing the work to supervision and as I said earlier, being the final decision maker.
• 5:42
  Tim Hardcastle: So underwriters, for example, can spend more time on exceptions, portfolio decisions, and relationships, and the work that they previously might have been caught up, which was stuck in their inbox and perhaps in spreadsheets and trying to chase topics around, that activity can be and should be passed to more of the AI capabilities so that the managers are spending more of their time on the high quality activities.
• 6:07
  Tim Hardcastle: And I think that's a good thing for MGAs and carriers.
• 6:10
  Tim Hardcastle: I think it's a good feeling for the industry because you know, often underwriters are most highly experienced, quite well remunerated.
• 6:17
  Tim Hardcastle: I think the quality of their input should be on the right things rather than, let's say, the slightly more rudimentary or admin related activities.
• 6:25
  Tim Hardcastle: So that's one thing.
• 6:26
  Tim Hardcastle: So the shift from doing the work to supervising, we're seeing, and I think we can expect more of.
• 6:32
  Tim Hardcastle: But I think the other interesting thing around jobs and the workplaces, and we're experiencing this ourselves as well, is new accountabilities are appearing.
• 6:41
  Tim Hardcastle: So we think of agents as like new employees.
• 6:45
  Tim Hardcastle: They need boundaries, job descriptions, a manager, a limit on what they can do and can't do.
• 6:51
  Tim Hardcastle: So someone has to own and manage the agents that are being deployed across an organization.
• 6:56
  Tim Hardcastle: So this is a new challenge for HR, for how team members interact with AI.
• 7:01
  Tim Hardcastle: And I think again, that's an emerging area.
• 7:04
  Tim Hardcastle: The one that perhaps worries me most for the industry is the apprenticeship problem, which is where junior people, you know, leaving university, whatever A levels coming into the industry, they need to learn by the more straightforward tasks.
• 7:18
  Tim Hardcastle: And obviously, AI can now do that.
• 7:21
  Tim Hardcastle: The question I think for many MGAs and carriers as they're building their businesses and taking the benefits of this more efficient way of working, is how do they find better routes for more junior members of the team to come through and learn how to make judgments and get the experience that traditionally the historical way of looking at a classic pyramid structure on an organization would have allowed them to start at a junior level and then move their way up.
• 7:46
  Tim Hardcastle: And that's obviously not so available now as it was literally six or nine months ago.
• 7:51
  Tim Hardcastle: So I think there's quite profound impacts on organization and roles within carries and MGAs.
• 7:58
  Abbie Day: I mean, the training point that you said is something that I'm talking to so many people about.
• 8:02
  Abbie Day: How do we completely train our staff to work with this new technology?
• 8:06
  Abbie Day: One thing I think about a lot is it feels a lot easier to check our own work that we've done than it is to check an AI's work in some cases.
• 8:16
  Abbie Day: What kind of training needs to be offered to new staff, to managers that are already in place?
• 8:22
  Abbie Day: Does it go beyond just the skills that we're using with AI into kind of new management styles and things like that?
• 8:28
  Tim Hardcastle: There's multiple layers to that point you're making.
• 8:31
  Tim Hardcastle: Whilst I've talked about the work shifting from trying to do all tasks more to this higher quality oversight, final making the judgment, I think in some roles, the skill development or the management of the activity by AI is quite straightforward because the underwriters typically that are in those positions, making those decisions on complex risks, they're very experienced.
• 8:53
  Tim Hardcastle: They know what right looks like.
• 8:56
  Tim Hardcastle: They'll use various different sources of information and other judgment to form and view.
• 9:00
  Tim Hardcastle: And I think for those kinds of groups and communities, I don't think there will be necessarily particularly retraining.
• 9:07
  Tim Hardcastle: I think it's like they'll be managing historically more like the junior underwriters, and now increasingly they'll be using or managing AI fulfilling some of the tasks of the junior underwriters.
• 9:17
  Tim Hardcastle: So I think they would just review and see that in the same way that they would have done with junior underwriter individuals.
• 9:23
  Tim Hardcastle: So I think for that group, it's straightforward.
• 9:26
  Tim Hardcastle: When you go to the other end of things, as we talked about earlier, with the human-enabled, where I was talking about AI working at scale, and then the humans are allowing the AI to do the work and they're actually judging the exceptions.
• 9:39
  Tim Hardcastle: That's where what we're seeing some organizations is doing is finding multiple ways of layers of governance.
• 9:45
  Tim Hardcastle: And we're doing the same thing with the way that we provide some of the AI support across our platform, where you'll have a mixture of a separation of governance from the AI that's actually doing the work, and then another layer of governance, which again is a separate AI capability, which is doing some of the reviewing.
• 10:01
  Tim Hardcastle: So you end up with some people are calling it a council, where you have different AI models reviewing each other's work, and you're triangulating then how the producer of the workflow, if you like, is then doing that to a level of accuracy because the humans are checking the exceptions, but they're also using AI to do the reviewing of the, let's call it the originator's work.
• 10:25
  Tim Hardcastle: And for skills that you need for that, that's starts to become a set of skills that people need to know about how to run, manage, and orchestrate and put controls around AI.
• 10:37
  Tim Hardcastle: That's an area where training and development is literally having to catch up in real time with the AI capabilities that are getting better and better.
• 10:45
  Tim Hardcastle: And so what we're seeing is across our client base or companies we're talking to that there are activities going on where people are being helped to put themselves in a position to fully understand the limits, the constraints on AI activities from this workflow, how they use AI to do the governance and the oversight on the AI itself.
• 11:04
  Tim Hardcastle: And then finally the human layer where they're looking at exceptions.
• 11:08
  Tim Hardcastle: The training and development that's going on is something that is happening in a very dynamic way because people are scrambling to put in formal learning mechanisms.
• 11:18
  Tim Hardcastle: Even if you go to, for example, you can see at the upper echelons of training that there are now master level courses being put in place with some of the leading universities in the US and the UK for AI strategy, AI governance, AI management.
• 11:32
  Tim Hardcastle: And some of the leaders that we talk to have been already been on some of these courses, which are being put in in an accelerated way, say three months.
• 11:39
  Tim Hardcastle: And the final point I'd make that come back to my earlier issue around how do you then nurture and develop talent as it comes through when you're enriching the organization by using more AI.
• 11:50
  Tim Hardcastle: That's the one I don't have an immediate answer for.
• 11:53
  Tim Hardcastle: I think companies are tackling that in different ways, but I think that is an industry-wide challenge and something that I'm, you know, all ears around about people that are coming up with new initiatives to achieve that.
• 12:03
  Abbie Day: And you mentioned earlier this kind of pyramid shape that we have in place for companies at the minute.
• 12:08
  Abbie Day: It sounds like from what you've been saying, that could look to change as AI comes in and we're managing people alongside AI.
• 12:16
  Abbie Day: What are you expecting that shape to look like, say, five years in the future?
• 12:21
  Tim Hardcastle: I think what we can foresee is that that classic pyramid is going to narrow at the bottom.
• 12:27
  Tim Hardcastle: And so there is an open question on where are tomorrow's senior underwriters going to come from?
• 12:32
  Tim Hardcastle: There's a challenge for the industry around the fact that, as I said earlier, that senior underwriters have built accumulated years of experience when they're managing AI that's supporting them on their decision making, they in a very comfortable space because the AI is removing some of the unnecessary work they were doing.
• 12:50
  Tim Hardcastle: And the question remains open is how do you bring people through to get up to that level of seniority?
• 12:56
  Tim Hardcastle: I think inevitably what we'll see is that there still will be processes.
• 13:01
  Tim Hardcastle: And we, you know, for example, I was um just talking to one of the main consulting firms the other day.
• 13:07
  Tim Hardcastle: They're still maintaining their graduate intake program, putting their graduates through structured training.
• 13:12
  Tim Hardcastle: But as I said earlier, that training is now orientated and skewed very heavily towards understanding how to manage, orchestrate, and govern AI.
• 13:20
  Tim Hardcastle: And I see companies doing the same thing, but I think we'd have to recognise from an economic standpoint that not as many graduates are going to be taken into the industry to fulfil tomorrow's underwriting roles as it would have been a year ago.
• 13:34
  Tim Hardcastle: I am concerned about that, but I think it's an inevitability of what we're seeing in terms of the displacement in the workplace of AI doing the jobs that previously would have been done by the juniors in the underwriting team.
• 13:46
  Abbie Day: The last question does ask you to look ahead slightly.
• 13:49
  Abbie Day: I'd like to know whether you think the insurance industry or the companies within it lend themselves to use cases where humans eventually aren't in the loop.
• 13:59
  Tim Hardcastle: If you take a step back and look at the context and the environment in which MGAs carriers, particularly that we know very well, and even uh and even brokers in the environment in which they operate.
• 14:09
  Tim Hardcastle: I can't see a future where we're going to have a world in which all transactions, all relationships, if you take the broker component, all underwriting decisions across all lines of business is fully automated.
• 14:22
  Tim Hardcastle: I think that is fanciful.
• 14:24
  Tim Hardcastle: What we have to recognize is the context within which brokers and carriers and MJ operated.
• 14:29
  Tim Hardcastle: They both work within regulated environments.
• 14:32
  Tim Hardcastle: They have responsibilities, not just to their customers, but obviously within that regulated context.
• 14:38
  Tim Hardcastle: And just because AI is available and is potentially a force for improved efficiency and streamlining operations, it doesn't remove the accountability for ultimately making decisions.
• 14:52
  Tim Hardcastle: And so in any context in which AI is being used, there are two or three questions that people need to be able to ask, which is what happens if it goes wrong?
• 15:01
  Tim Hardcastle: Can we validate, determine, and be explicit about how decisions were made?
• 15:05
  Tim Hardcastle: Because you still, whether you're a carrier or an MGA, ultimately you're if decisions are made by AI and go badly wrong, you're accountable to the regulator for those decisions.
• 15:17
  Tim Hardcastle: So I can't see in any scenario in the near term, and I'm talking three to five years, where you're going to move to a fully automated environment.
• 15:25
  Tim Hardcastle: There will be improvements in, as I said, the underwriting operations and other parts of the insurance value chain where, you know, ingestion, data analysis, these things can be done by AI.
• 15:36
  Tim Hardcastle: But I can't see a future where humans, whether that's human-led, human-in-the-loop, or human-enabled, will be removed completely.
• 15:42
  Abbie Day: So it's not about to take all our jobs then?
• 15:44
  Tim Hardcastle: No, I think we can be confident that there will be plenty of employment.
• 15:48
  Tim Hardcastle: And I guess the other perspective to end on a slightly more positive note, because we were talking about potentially some of the pyramid collapsing for some of the junior intake.
• 15:57
  Tim Hardcastle: You know, let's take a step back and think about the world of insurance in terms of what's the opportunity for the insurance community and insurance players.
• 16:06
  Tim Hardcastle: And I would say this that there is a huge unmet need of insurance.
• 16:12
  Tim Hardcastle: And that crosses the span of PC and life and health.
• 16:17
  Tim Hardcastle: So there's nowhere near enough people who are having life insurance, health insurance, there's nowhere near enough businesses that are fully covered from an insurance standpoint on all of their assets and the services that they deliver.
• 16:28
  Tim Hardcastle: And so there's a huge unmet need across all lines of insurance, I would argue.
• 16:32
  Tim Hardcastle: And the opportunity is for all of the MGAs and carriers, while they can lean out and be more streamlined and become more efficient because of AI, they can also look to grow businesses by serving this unmet need.
• 16:47
  Tim Hardcastle: And then that should, everything else being equal, bring in far more people then to be supporting that growth.
• 16:54
  Tim Hardcastle: Because notwithstanding there's going to be a reduction in some of the activities that maybe juniors would have undertaken.
• 17:00
  Tim Hardcastle: If the businesses can be growing to fulfill these unmet needs, then in large part, timing is always an issue about what happens first.
• 17:08
  Tim Hardcastle: But in large part, I think you could see the industry certainly being confident that it should be employing the same, if not more people in that same time horizon over three to five years, because again, they can get more efficient, they can serve these unmet needs more effectively.
• 17:24
  Tim Hardcastle: And therefore, there's a big growth agenda, I think, for many companies to look forward to.
• 17:29
  Abbie Day: Amazing.
• 17:30
  Abbie Day: I think that's a good note to finish on.
• 17:31
  Abbie Day: So, Tim, thank you so much for chatting with me today.
• 17:34
  Abbie Day: It's been really, really interesting.
• 17:36
  Abbie Day: Thank you.
• 17:36
  Tim Hardcastle: Pleasure