E314 - Reluctantly, another episode on AI

(A) E314 - Reluctantly, another episode on AI

===

[00:00:00]

[00:00:00] Nick Roome: Hello everybody. Welcome back to another episode of Human Factors Cast. This is episode 314. Can you believe we've done that? Many of these? We're recording this episode live on February 5th, 2026. I'm your host, Nick Rome, and I'm joined today by Mr. Barry Kirby.

[00:00:17] Barry Kirby: Hello there. How are we doing?

[00:00:18] Nick Roome: Do, doing all right, Barry, how are you?

[00:00:21] Barry Kirby: I'm very well, thank you. Loving, love to be here as always.

[00:00:24] Nick Roome: Me too. I've turned off the pre-show apathy and have turned on the show. Persona, we're here. Uh, we have an awesome show lined up for you all. Let's talk about the title. I'm begrudgingly accepted that we must talk about AI this week as much as we try to avoid it, as much as we try to get you interesting stories that do not involve AI and human factors, well, sometimes it's just unavoidable.

[00:00:47] Nick Roome: So we'll be, we'll be previewing a few different stories about what the latest in AI is. But first we got some programming notes. In fact if you're unfamiliar, we have a lab. We have a lab here at the Human [00:01:00] Factors cast, that's the human factors cast. Digital Media lab rolls right off the tongue, and we have some fabulous folks in that lab and what those fabulous folks are working on sometimes see the light of day.

[00:01:10] Nick Roome: Sometimes it's the unseen, unsung appreciation for behind the scenes tasking that happens there too. But we have something new for you all that is a brand new take on human factors and its crossover with games. And so it is called user experience points. This is a very clever title.

[00:01:31] Nick Roome: We're very proud of it. And, uh, you know, the first episode is out now, so if you follow us on any of the social platforms that have short form video, you can find it there. YouTube, TikTok, Instagram, uh, it's a great, it's a great reel. It's a great it's a great time and we're gonna be posting more of those.

[00:01:46] Nick Roome: We're heavily into development already, and, uh, it's been doing some numbers on, on the Instagrams. It's, I think, what, 2020 something thousand views, uh, which is not insignificant for, uh, niche like human [00:02:00] factors. Very cool. Great job lab. Aside from that, Barry, what's the latest over a 1202?

[00:02:05] Barry Kirby: 1202.

[00:02:05] Barry Kirby: We actually have some live episodes or show, I say a live episode. So the first episode since 2024, um, is now live out there. So Matt from Mil UX shared his thoughts with me on how UX and HF can be better used and better recognized in the defense domain. And really, we spent a lot of time talking about, um, how do we better appreciate the value that it brings to defense products?

[00:02:30] Barry Kirby: But we also went back to that, um, that the old argument, the old discussion around what is that relationship between HF and ux? And Matt comes from this as a, as a rare, comparatively newer into the domain. So he's, he's ex-military. And then picked this up on a project and has since become almost a a evangelical about it in terms of using, uh, UX and Kanban and them sort of approaches.

[00:02:55] Barry Kirby: So we had a really good chat around that. So that is out there. I've also recorded the second episode, so I'm not gonna share who [00:03:00] it's with you yet, but it's somebody who's quite important and that's gonna be live, going live in a couple of weeks. And then just a reminder that if you've been using our website to access 1202, then obviously we have the slight hiccup with the old, uh, 1202 podcast domain.

[00:03:14] Barry Kirby: Um, so it's now 1202 pod.com. So see you there.

[00:03:18] Nick Roome: You know, as long as we're talking about new stuff, we have so many new stuff for you this week. We got the user experience points. We got the new stuff over at 1202 and local to the show. We have a new segment for y'all tonight. So I'm very excited for y'all to see the new segment that, uh, that the aerospace systems technical group from HFES has been cooking up for us.

[00:03:37] Nick Roome: So, uh, that'll be a little bit later in the show, but why don't we go ahead and get into the news.

[00:03:43] Nick Roome: That's right. This is the part of the show all about human factors news. Barry, what are our stories? This, or let's start with one story. What is our story? Top story? What is the first story this week? Okay, just three story go

[00:03:58] Barry Kirby: as you say. I have [00:04:00] three stories this week. So the first one, as we are gonna look at is what happens when novelty and with, and it starts quietly rewiring how we think, how we work, and how we play.

[00:04:11] Barry Kirby: Our first story starts tackles ai slop and how it may actually be training us to act more like Wikipedia editors. So we, we are now more into double checking sources and treating everything like we see online with default skepticism instead of blind trust. Nick, have you been finding the, do you check your sources more?

[00:04:31] Barry Kirby: Do you believe what you read?

[00:04:32] Nick Roome: It's, well, let me, let me talk about it from this perspective. I think this this claim here is kind of optimistic at best. Uh, I, I tend to think about like the folks on Facebook, the Facebook moms that are like, oh my gosh, can you, how did they train those cats to dance?

[00:04:49] Nick Roome: You know, I, I tend to think about those people as the average, where there are people like us who are heavily invested in the advancements in AI and the advancements in [00:05:00] tech. And I think that we are certainly changing our behaviors. Um, and I think that there's also just an interesting note that if we do sort of default to this skepticism of what we're seeing, I think that can tend to breed conspiracy theories, which is incredibly dangerous.

[00:05:21] Nick Roome: And so those are my two, like initial thoughts, but we can talk about it more in detail. What, what were your thoughts as you were reading through this article?

[00:05:28] Barry Kirby: So, I guess in some ways, similar to yourself, that it's easy to see some of this data coming out and thinking, yes, people are reading deeper into it and that type of thing.

[00:05:38] Barry Kirby: But as you point up, is this just a we, we, we've, we're seeing our own stuff on social media, so we believe what we see and therefore if we see more people doing this sort of thing. We, we think we can transcribe that to the whole population. It's not quite true. I think also though there is still some elements within the technology that means that, you know, [00:06:00] for those of who, when we are reading it it's, it's very analogous to what, to simulation and what we see in simulation.

[00:06:05] Barry Kirby: Because there is this term uncanny valley in simulation where something is simulated well, but it's just not quite right. You know, there, it, it, it isn't perfectly simulated. So you in your brain can know it looks good, but it's not. But it's, and it looked wait high quality and all that sort of stuff, but it's just not moving in the right way.

[00:06:23] Barry Kirby: The, the, the creature or the person is not moving. It's, but we push ourselves because we're like, okay, I know it's not real, but I'll believe it anyway in order for to, for the simulation, for the gaming, for whatever it is that I'm doing. And I think that's the way AI is now, that you can see a lot of AI content, um, AI generated content through LMS or whatever.

[00:06:42] Barry Kirby: And for those of us who read quite a lot of it, we can see it, we know it's AI generated because of the construction. Um, and because of what it's put together. The, the video is, is just that either a little bit too perfect or a little bit too polished. But we, um, but as I've said, 'cause that's in our [00:07:00] bubble, the general public.

[00:07:02] Barry Kirby: It's possibly not quite there, and we're still in that piece of believing what we see is the, it's still got a fair way to go before we, everybody is as critical as they need to be. But it is this, this idea about being a be critically reviewing our sources is a key skill we're going to need in the future.

[00:07:21] Barry Kirby: It's easy. I, I, I sort of make the analogy with newspapers of old, we sort of talk about this idea about lies in the press or nuances in the press that, um, lean left or right or whatever, like it's a brand new thing. It's not, it's always been in the press, but we've had it through dis discrete channels before and you've known where a certain newspaper, what it's leanings are, whereas now we, you don't have them big flags up there.

[00:07:45] Barry Kirby: So yeah, the, it's gonna be a key skill. I do think it is becoming more, and the data seems to be backing that up, but it's not everybody, as you quite rightly say.

[00:07:52] Nick Roome: Yeah, I think there's also some things going on here when it comes to AI slop. Especially when it comes to the [00:08:00] assumption. I think there's an assumption by the author of this one particularly that is thinking that people are ramping up their literacy, which when you know, I, I kind of see the opposite.

[00:08:12] Nick Roome: People are disengaging or they're sort of retreating into their echo chambers. And so I, I, I questioned the core hypothesis around this article. They do cite some data that would support their claim, right? So I'll, I'll just read some of these, uh, supporting data points here. There's 21% of videos shown to new YouTube users are classified as low quality AI slop.

[00:08:36] Nick Roome: So there's, that's kind of the baseline for what normal content looks like for a new account. You also have a, another citation here for the trust in national news dropping 76% in 2016 to 56% in 2025. That's a 20% reduction. And I, I think, um, they, they make the claim that there's [00:09:00] this, uh, this might be fake by default thing happening, but, um, I don't know if that's necessarily true, especially when.

[00:09:08] Nick Roome: The, the contrary is true too. AI generation, AI video generation is getting better and better, and it's harder for us to discern that uncanny valley of what is fake and what is real. The tells are getting harder and harder to parse apart. There are, there have been videos that I've seen that have, that I've said like, please be ai because they're, they're so disturbing or something, you know, that it's like, Hey, that's gotta be ai, right?

[00:09:34] Nick Roome: Or it's somebody that's deeply disturbed doing these things. So I don't know. I I, I tend to err on the side of this one is I see it, but I don't.

[00:09:44] Barry Kirby: Yeah, I think the, the one statistic in this, I think is true, is really interesting from a humanity perspective is the 70% of adults, young adults get news incidentally.

[00:09:56] Barry Kirby: So they are coming across news. They are not going out to try and [00:10:00] find news where historically you've gone by the newspaper, you go and look and work out what's going on. In the new in the world today we, they are now, we, we now have to put it in front of people more rather than them coming to Graviton and consume it.

[00:10:15] Barry Kirby: Um, the, the bit that, the other bit in this article I quite liked is, which I hadn't really heard of much about before, is this C two pa, which is a, a way a, a tech solution for Crip cryptographic signing. So to prove that a file is to, to prove that it's real. So it's real content. It's, um, there, now it's only really viable, um, for a handful of devices this article says.

[00:10:36] Barry Kirby: Um, so unless you make it universally available, um, which could then, you know, how do, how does it not then get abused?

[00:10:43] Nick Roome: Yeah. I

[00:10:44] Barry Kirby: I, the idea of being able to sign, you know, this content today is we are to still live people doing our thing across the outline. Far

we

[00:10:54] Nick Roome: know,

[00:10:54] Barry Kirby: One day we, we've, where we sort of joked about it before, isn't it?

[00:10:58] Barry Kirby: Where we've said, said, can we have two AI [00:11:00] versions of ourselves doing a podcast on our behalf? Um, that's not today. We are still here, we're still human, but we would need, at some point, we are probably gonna need to sign, uh, digitally sign in some way to say, yes, we are real. So I quite like that, that that was a, that was a a, a good insight.

[00:11:15] Nick Roome: Do do I need to, do I need to add like a, a ticker at the bottom that says we are real people talking about ai?

[00:11:24] Barry Kirby: Yes. Yeah, because it's a reversal and then because then we'll have the, um, the AI versions of ourselves talking about the, uh, talking about the real people.

[00:11:32] Nick Roome: Okay. Why don't we move on to the next story here.

[00:11:35] Barry Kirby: Okay. So the next story, we are digging into MIT Technology reviews. Look at the augmented workplace where AI shows up as a general purpose tech technology that raises productivity by amplifying human work rather than replacing whole jobs, especially in under automated sectors like healthcare, education, finance.

[00:11:55] Barry Kirby: I probably took research in there as well. Um. Yeah, [00:12:00] do, do you see, um, have your AI buddy alongside you, um, helping you in your day to day, uh, rather than taking over your entire job?

[00:12:08] Nick Roome: In my notes, I put the human AI robot teaming technical group exists for this very reason. I think there's a large consensus within our field that that it's going to be a teaming approach, that AI is not to be seen as a tool, but rather as a collaborative a collaborative partner in, in various circumstances.

[00:12:29] Nick Roome: And I think the more that we think about that as, you know, this AI thing is a partner, this AI thing, AI as a partner, especially agents or even LLMs, I think there's a, a different mindset that, that comes to that approach than thinking about them as tools. And so I think this article is kind of getting at that, and it'll be interesting to dig into this a little bit deeper and what does that actually mean for the workplaces.

[00:12:58] Nick Roome: But what are your initial [00:13:00] thoughts on this one, Barry?

[00:13:01] Barry Kirby: So for me, I mean, this is MO'S law writ large. So the idea of technology evolving and developing a, a rapid pace, ai, we spoke about it in almost residential terms. I remember we started studying it, what, in 97, 98 at university. And it was, you know, it was a real tight thing that you have to be, super into software and all that sort of stuff to, to, to develop this stuff.

[00:13:25] Barry Kirby: But now it's so embedded in what we're doing. It's, it's, it is rapidly changing the way we work, or it should be rapidly changing the way we work because of what it's doing. So you're right, the, uh, it's gonna be working alongside us, but there are them direct tools. Um, we, we've talked about things like vibe, coding, you know, that changing the way software engineers work.

[00:13:44] Barry Kirby: The companion. So the, what this story largely links into with this idea of supporting what you're doing it being alongside to nudge you and give you information to your fingertips. But it's even there in indirect support. So a lot of your web searches and that now is [00:14:00] supported by large language models to give you, to make that searching easier.

[00:14:04] Barry Kirby: So you now, if you're into Google or any of the other things, you're now getting the AI prey at the top. So that's sort of, you know, that's giving you the summary of your research to, so you don't even have to scroll through your things anymore. I think for me, the big step that is, is really happening now is.

[00:14:22] Barry Kirby: The gatekeepers have changed. So really up to this point, I would say up up to the past, past few years, the specialists who can use this technology, who can define this technology, who can code in this technology have been the gatekeepers to the technology. And in, and that's in the same way as that, you know, the data itself was was, uh, gate kept by librarians and all that, all, you know, them and companies and large organizations until the internet came along and made in data.

[00:14:53] Barry Kirby: So, um, so available to the point now we, we treat it almost like a currency. So, that the internet [00:15:00] itself heralded that data revolution, and now we, AI is clearly a the next revolution, um, that's gonna enable so many things. Uh, I just, I think there's a lot of people out there and some people I really respect on the likes of LinkedIn and places like that who are still trying to put the AI cat back in the bag.

[00:15:20] Barry Kirby: They're trying to sort of close Pandora's box and it's just, I just don't, that it's just not an option anymore. So as you've alluded to, there's for the HF domain itself, there's kind of, there's three mainstreams I see that we need to get good in it and it's not an option. It has gotta be part of our our everyday work.

[00:15:39] Barry Kirby: And that's understanding trust. And where data is coming from, understanding teaming and task allocation that derivation between automation, autonomy, and human work, um, is they're gonna be three key pillars of, of work for us going forward. So yeah, I'm quite, I'm, I'm quite excited by this.

[00:15:59] Barry Kirby: Having worked [00:16:00] on projects, I think probably about eight years ago, um, which I've mentioned on here before about, um, using, um, so there was n avatar, uh, made of myself to do a representation of mean, to look to see whether we could, um, if a chatbot version of me was as good as the real me. Now this seems like it's just a thing.

[00:16:20] Barry Kirby: It's it's there.

[00:16:22] Nick Roome: Yeah. One, one thing that's particularly striking to me about this article MIT Tech Review is an outlet for people who love nerdy stuff and, which is why, you know, we find a lot of our human factors stories in there. And the interesting thing to me about this particular one is that this is in partnership with Vanguard.

[00:16:42] Nick Roome: Uh, Vanguard is not small by any means. And the thing that's striking to me particularly about this is that we as human factors practitioners, we've known this for a while, right? We've, we've known the teaming approach, we've known the, the domain as far as it approaches [00:17:00] things like decision making and overloading cognitive tasking and, and those types of things.

[00:17:06] Nick Roome: But the interesting thing to me about this is that now you have have these big companies kind of stepping in and saying, maybe we should shift away from replacing jobs to, or automating jobs, to working alongside this thing. And so that to me is a distinct difference or a distinct marker that if you can get big businesses to approach this perspective, then our job becomes easier to justify and easier to promote, and easier to talk about, because now we're understanding it as a universal need rather than thinking about it as, we gotta replace these humans with these agents or something, you know, versus, okay, we gotta get these humans aligned with these agents.

[00:17:48] Nick Roome: It's a very different problem, a very different mindset, very different space. So that's very interesting to me. And I'm, I'm happy to see it and I'm, I'm, I'm genuinely thrilled for our folks, uh, in the [00:18:00] human factors field who are gonna have a ton of work to do in this space. Mm-hmm.

[00:18:04] Barry Kirby: I guess the, the one quote outta this that I think is absolutely spot on.

[00:18:08] Barry Kirby: So part of this framework, this review has allowed them to ex examine auto, uh, AI automation risks over 800 different op occupations, uh, which is found. So they've done a really broad review of this. And fundamentally, they, they, they reckon they think that upwards of 20% of occupations could be at risk.

[00:18:25] Barry Kirby: That's, which is actually a lot smaller than normal than what we talked about in the past. But what they say in that quoting from the article is, Workers' time will increasingly shift to higher value and uniquely human tasks, which for me is all around making decision, making, moralistic decision making, uh, value decision making.

[00:18:45] Barry Kirby: But you the, that hu unique human thing is to make decisions. Yes, AI helps us with the data behind that and things like that. So I'm really pleased to see that in articles like this we're now really talking about the, what that, what is that u what are the uniquely human [00:19:00] traits that AI just will not be able to do or shouldn't do?

[00:19:04] Barry Kirby: Uh, where, where, where is that human in the box? So as you say, it's not just about replacing which I think is fab.

[00:19:10] Nick Roome: You know, one, one thing I think. Generally that businesses should be cautious of as they go into this approach is, you're right. You know, there's gonna be these distinctly human tasks, but there are also other things here at stake that the people who are changing their mindset need to consider, especially when it comes to thinking about their employees as like co-designers of the workflows or the processes that have to happen as you interact with this agent.

[00:19:40] Nick Roome: Uh, they, they can't be prescriptive about it. They need to include folks in on this process. So that way when it is applied, it's not applied in a way that is gonna be clunky or where sort of it's tacked on. And I think that's the difference between thinking about it as a tool and thinking about it as a partner, [00:20:00] where if it's a partner, it's, it's got, it has to be more integrated into the workflow of the worker, regardless of the industry, right?

[00:20:09] Nick Roome: It has to be integrated into that workflow. Whereas if you had, uh, sort of that prescriptive approach approach, you must use AI to do X, y and ZI feel like that partnership breaks a little bit. And so that's something that I'm concerned about, looking at it from a human factors perspective, is to make sure that leadership or higher ups understand that it needs to be, you know, it needs to be thought of as, as a collaborative approach for how to move forward with these rather than a prescriptive one.

[00:20:38] Barry Kirby: Yeah. And I guess there's one of the challenges there is what type of AI do you then start developing? Because you, you think that if you're gonna have a cooperative AI agent, something to help you do your job, then that's gonna be, you know, as we have now, the AI are generally developed quite specifically to do specific tasks or specific bouts of [00:21:00] knowledge.

[00:21:00] Barry Kirby: Whereas artificial general intelligence is the thing that's coming up. And is it easier then for us just to produce a GI agents that will support any job? And then does that allow, is that then gonna, gonna support the user? So on the one hand it will answer any questions or support us in any task that we need to do from the specific job to help him make a coffee or something like that.

[00:21:25] Barry Kirby: Or working out if your shop's open does that mean that then the AI agents can overstep boundaries of the jobs and things like that? So how, how is it cheaper to buy a bunch of the same a GI and just deploy that no matter what the role in the workspace and therefore, how do you delineate between them?

[00:21:42] Barry Kirby: And then I see, I, I do keep on going back to, and I, I, we've quoted this on the, on the show before was the quote from Professor Paul Salmon who, when he was talking about agis. And as they become more intelligent, more, the more that they know, how do we still monitor that knowledge base and understand [00:22:00] where that is in in overall intelligence, I guess.

[00:22:04] Barry Kirby: Because what does the future hold there? That is still, i, i having, being quite positive around this. I think, you know, it is the future. But what, where we still need to be cognizant of those risks. Within the workspace.

[00:22:18] Nick Roome: You bring up an interesting point that I don't know if we've talked about before, but essentially this boils down to, you know, if you think about the difference between a targeted model that is meant specifically for a specific task and you, you work on integrating that as a partner in the domain versus applying something like a general artificial intelligence, there's an interesting gap that happens.

[00:22:44] Nick Roome: I feel like the companies that are going to be able to afford the highly specific tailored models are going to far outperform the ones that have a general intelligence applied to many different things. And likewise, I feel that there's also going to be a, an [00:23:00] even bigger gap from the companies that choose to use AI as a strategy versus those who don't.

[00:23:06] Nick Roome: And so you'll get sort of this even wider gap between pe the companies who don't choose to incorporate AI into their employees workflows, all the way up to the companies that are using very targeted models for very targeted tasks in their domains. Right? I'm thinking about like healthcare for example.

[00:23:25] Nick Roome: You have very targeted models that are looking at you know, radiology images or something like that, right? That's a very targeted model. You don't want general intelligence looking at that stuff. Because the, the healthcare companies that. Do the general intelligence versus the targeted model, there's gonna be a wide difference in who, who is able to look at early screenings for things like cancer, right?

[00:23:48] Nick Roome: So that's just one example of that gap.

[00:23:51] Barry Kirby: Though,

[00:23:52] Nick Roome: yeah,

[00:23:52] Barry Kirby: sorry. I guess to jump in, I guess the, the only reason that the gap is gonna exist is because of processing power and an [00:24:00] economy because the, that targeted piece within the overall A GI gap, it just means that there's a lot of if, if they do incorporate it there, that that means there's a lot of fat on that model that's not going to get used in exercise.

[00:24:13] Barry Kirby: So if you can afford to run it both power wise and however that mechanism works, then maybe, then again the large companies will be able to afford to do it. And I do think there is an, there's an interesting piece now. I think the, particularly for small companies, micro companies, small companies, medium sized companies, that if they get on the AI bandwagon now they are, they will be still, they can still be on that crest of getting the AI advantage.

[00:24:42] Barry Kirby: Um, so if you are into developing, you know, you won't develop tools and stuff like that, now you, you've, there's still leverage to be had by, by adopting an AI strategy and making that work. We're gonna get to a very to a tipping point, I think pretty soon, where if you are not using AI as part of your business, [00:25:00] you are gonna be left behind.

[00:25:02] Barry Kirby: You know, and, and that, that differentiator, I think we are still on the, the, um, the, the, the growth side of it, but only just if you, even a small, so maybe an electrician or something like that. So maybe somebody who isn't into software into that type of thing, but still has, uh, a lot of use for AI to help streamline their business flows.

[00:25:22] Barry Kirby: If they're not using it soon, then they're gonna get left behind, um, by other people who do adopt. Yeah, it's gonna be interesting to see where this narrative shifts from the could have and it's a great thing to do, to must have, and if you don't you're gonna fall by the wayside.

[00:25:37] Nick Roome: Yeah, and I, I think ultimately it comes down to leadership's attitudes on how to implement it, right?

[00:25:43] Nick Roome: If they think about it as a tool and they are prescriptive about it, versus if they use specialized models versus if they just brute force it, that all that stuff will impact how it's adopted within a company. I, I'm just happy to see that, that a big company is, is waving the [00:26:00] flag on this.

[00:26:00] Nick Roome: That's, that's my takeaway.

[00:26:02] Barry Kirby: I guess the other bit that I would say as well is that this is we are at this point now where we don't, we probably don't know where AI's truly going to go because it's now got to the point where the curious can use it. So it's the likes of you and I sitting there going, oh.

[00:26:20] Barry Kirby: What if I could do this with it and what I could, I've had this idea, could I just do this? Could I link this to this, uh, this agent, to this agent, and do that? And some people will come up with some quite wacky ideas or what we consider wacky and then just prove things that nobody was expecting was on nobody's root map or anything like that.

[00:26:39] Barry Kirby: And then we'll end up using them on a day-to-day basis. I mean, could we say that five, 10 years ago that we'd be using AI so much. So even just within this, within this podcast, the amount of tools and things that you've got, we've got in place to make this a better production so we can focus on the delivery rather than having to do what would take you hours and hours and hours of prep [00:27:00] beforehand.

[00:27:00] Barry Kirby: We wouldn't have predicted that years ago 'cause it would've been seen as a frivolous use of this type of thing.

[00:27:04] Phil Doyon: Mm-hmm.

[00:27:05] Barry Kirby: Uh, whereas now it's a core thing of what we're doing. So, yeah. No, I I think they, this idea, it's, this is the, almost the Pandora back Pandora's box moment where people are, um, are gonna be using it for doing random things that we, we can't even think about.

[00:27:19] Barry Kirby: And then it will become a defacto standard in very rapidly

[00:27:24] Nick Roome: open the box. All right. Uh, why don't we get into the last story we have up tonight.

[00:27:32] Barry Kirby: Yeah, so we're gonna go with Google. So we play with Google's Deep Minds Project Genie, um, a showcase for the Genie three world model that lets you sketch or prompt your own tiny interactive worlds.

[00:27:47] Barry Kirby: Nick, have you been into your, is it 62nd version of this yet? Have you handed over the, the, the big dollars to make this a thing?

[00:27:54] Nick Roome: Oh, no, not yet. No. It, it's, it's insanely cost prohibitive. I'll start with that. But I think [00:28:00] the, the more interesting thing about this particular, so if you're unfamiliar, what it does is it basically generate, it generates an entire environment and responds to your controls within that environment, right?

[00:28:14] Nick Roome: So you can think about like a video game, but a procedurally generated video game that only it generates every frame as it responds to your input. So that's, that's kind of what we're talking about here. And I think this is an interesting application outside of gaming, right? There's a lot of, uh, there's a lot of talk in the gaming spheres about like, oh, this is gonna, this is gonna cause developers to lose all their money, or, I don't know.

[00:28:39] Nick Roome: But, but I think there's a more interesting narrative here that, that maybe hasn't been talked about. And that's sort of the shift towards things like micro simulations. So now you might be able to you might be able to simulate how a a new tool in an environment might perform or how, a setup at a desk [00:29:00] might impact somebody's workflow as they're doing a, a certain task.

[00:29:05] Nick Roome: And it, it can be modeled live on the fly with something like this. And so I, I find that approach fascinating. I also find the applications to video games fascinating too. You can put whatever world you want in there and then make your own game. I think there's a lot more that goes into game development that will make this not a non-issue for quite some time.

[00:29:26] Nick Roome: But I don't know. What, what are your, what are your thoughts on this thing?

[00:29:30] Barry Kirby: I guess in some ways I'm, I'm, I'm not quite as excited about this as I, as I think I should be. It's really got interesting potential. Clearly it, it's, you watch the video and I recommend everybody go and have a look at the vid, have a look at the video because it's it's high quality.

[00:29:44] Barry Kirby: It's got, the way that they describe it, it's quite cool, but it's a bit, it's a bit Googled glass, you know, they produce something. It looks cool but what's it for really? What, what is the intended use case? I mean, my first thought was actually, what is this? Just Minecraft [00:30:00] in, in high def. Um, but it just doesn't have the rest of it.

[00:30:05] Barry Kirby: Or is it second life? You know? 'cause if those who remember Second Life as the next big thing without friends, but with a massive bill. 'cause as we've alluded to, it's, you know, you have to be you have to have their $250 a month AI Ultra plan to engage with it at the moment that'll come down.

[00:30:21] Barry Kirby: I'm, you know, this is clearly as it as with the done with Google Glass. It's an experimental thing. It's to throw it out there and just see kinda what happens. Um, so that bill will come down, um, in time. And I'm assuming that time limited factor is just for performance. You know, they can't be throwing that much.

[00:30:38] Barry Kirby: Well is clearly high end processing to everybody be just 'cause they're mini expense. I guess the use cases is, you've already touched on a couple. I see, you know, mini test rigs. Is there, I sort of had the idea. So when, when you look at digital twins, could you see a, and you weren't sure about how something was going to, um.

[00:30:59] Barry Kirby: H [00:31:00] working gravity or maybe not, maybe zero gravity or something like that. Understand the mechanical or a movement of something when you do an assessment for usability or, or something like that. I think there's a, there, there's, I can see different use cases in the same way that we Googled last. I could see when wearing them, when using them, or I could use it for this, I could use it for this.

[00:31:17] Barry Kirby: Surgeons are there good, oh, we could use it for this. But, um, I think it's gotta allow pe people are gonna be able to play with it and then see what they do for, then Google will do what they did last with Google last and go, okay, now we'll take it back in. And we'll go down this route with it.

[00:31:33] Barry Kirby: Yeah, I go, unlike the other two bits, it's a, it's a, it's a, it is a new part because obviously it is prompt engineered. Um, and so they're really leaning into that quite a lot. Um, so that is called, it's, it's, it's a great application of it. I can't quite get as, get as excited by it as the other bits.

[00:31:50] Nick Roome: You know, you, you bring up a good point about killed by Google, right? And I think this is actually a direct result of something that was killed by Google. I was a big, big fan of [00:32:00] Stadia when it was a thing. And one of the things that they were working on with Stadia was if you don't remember, stadia was a cloud gaming platform by Google.

[00:32:10] Nick Roome: And one of the things that they were working on in the early development of that was doing predictive frames based on predicted input from what you thought or from what the system thought your next input would be. So it was generating different frames behind the scenes. I probably similar to how this works to where if you were to hit left on the controller, it would generate, you know, these, these things and then it, the system would catch up with it.

[00:32:39] Nick Roome: And so that was an, an attempt to reduce the lag that people would experience on these cloud gaming platforms. So I think this is probably a lot of that research being realized now in terms of, of being able to produce not only frames ahead, but an entire model of a world at a prompt level.

[00:32:58] Nick Roome: And I, you know, I, [00:33:00] I tend to share. Your sentiment about enthusiasm for Google made products when there's not maybe some super strong use cases for it, but if you think about things like Google Glass, they're coming out with the glasses later this year, right? And so all the early research that they did on Google Glass paved the way for them to do something like that.

[00:33:20] Nick Roome: And even if, and, and Stadia again, killed by Google, but paved the way for stuff like this, I'm pretty sure. And then, you know, if this thing does get killed, what's the next thing that's going to use it? Yeah. So I see, I see a lot of value in this and I, I already kind of mentioned the simulation side of things.

[00:33:38] Nick Roome: I think that's a huge benefit to this technology outside of the gaming application, right? Because it's, you can imagine whatever world, but if you can imagine a real world, and if it, if it moved with similar physics or, you know, interacted with physics in a realistic way, you could take a picture of an environment or, throw in a modeled environment [00:34:00] and have humans interact with it in a way in a digital environment in ways that, that they would be anticipated to do so in real life.

[00:34:08] Nick Roome: And so, to me, the simulation piece is big. Um. I am excited about this, so I will be excited for both of us.

[00:34:18] Barry Kirby: Okay. No, that, I mean, you are right. I mean, this is why I, I don't mention the Google glass in a, uh, Google Glass journey in a derogatory way. I think that, I think a lot of people do. It's because you are Right.

[00:34:29] Barry Kirby: You know, it's, it, they've developed a way of allowing, of basically bringing in, uh, people paying to be beta testers, you know, and, and it's, so that's, that they've got the, their beta model. They'll sell it, we'll engage with it, give them the feedback, and then they'll take that away and go do something else.

[00:34:46] Barry Kirby: And that, I think this'll be there. I think the, I do like the idea the more that we've discussed about it around that digital twin world, because I've done, and I'm still doing, um, using human, uh, digital modeling [00:35:00] to try different things. But it's always it's a, it's an event. I have to go to a special room where special gear, if you could use something like this on your desktop or just putting on just a standard headset.

[00:35:12] Barry Kirby: Um, so there doesn't need like loads and loads of rendering loads and loads of messy market. It is just, it's done for you. That's fantastic. And especially with Industry 5.0 coming along. So this idea of better, of, of the more human focused in industry development in and production, um, this could have a really big impact into that.

[00:35:32] Barry Kirby: So yeah, I guess I'll, I'll take, I, I, I guess I'll take back. I'm not as excited or I'm not excited about it. I'm perhaps just not as excited. So I've got, I'm, I'm a bit excited.

[00:35:42] Nick Roome: Okay. Alright. I'll, I'll take a bit excited. You know, I, I think the one thing that I'll end on with this is just imagine the, the domains that this could be applicable in.

[00:35:53] Nick Roome: I saw a presentation years ago on sort of a, a simulated environment for being able to conduct triage in a [00:36:00] battlefield, uh, scenario. And so, taking things like healthcare or even defense or heck you could do you, you could even do education for something like this. Imagine, you know, building a, a simulated environment of a solar system or something for a child, right?

[00:36:18] Nick Roome: Like, and having them interact with that. Like, my son's really into space right now. He would love to be able to explore space, you know, with his own character using a model like this. And I just, I, I think of how great that would be for things like education and, you know, I, I think that's cool, but, uh, be cool to see it in all these other, all these other areas too.

[00:36:39] Barry Kirby: No, absolutely.

[00:36:41] Nick Roome: All right. With that, I want to thank our friends over at In Gadget and MIT Tech Review and the UX Collective for our news stories. This week we post all the links to the original articles in our Discord where you can join us for more discussion on these stories and much more.

[00:36:59] Nick Roome: We're gonna take a [00:37:00] quick break, and then we're gonna see what's going on in aerospace. We'll be back right after this

[00:37:09] Nick Roome: Yes, huge. Thank you as always to our patrons, I see you, I hear you. You are supporting us in more ways than you can possibly imagine because you are also supporting the lab who does some fantastic things and you support all the tools that we use the web hosting domains, there's a ton of stuff that goes into running a podcast and a lab if you, uh, if you could believe it.

[00:37:32] Nick Roome: So thank you genuinely for all your support and if you have the financial means to do so, we would love your support as well. Uh, I don't really have a good segue for this, but we are gonna get into a new segment for the show. I'm very excited about this. This new segment is called This Week in Aerospace, and, uh, this is brought to you by our friends over at the HFES, aerospace Systems Technical Group.

[00:37:55] Nick Roome: So, Elena, Phil, over to you. What's the latest in [00:38:00] aerospace?

[00:38:00] Elena Zheng: Well, thanks Nick and Barry. This is Elena Zang, joined by Phil Doyle for the Aerospace TG at HIVS. We're joining the show to share the latest news in aerospace.

[00:38:13] Phil Doyon: So just before the holidays, the US Department of Transportation unveiled the first advanced air mobility roadmap or a a m to accelerate aerospace transformation.

[00:38:25] Phil Doyon: That's a major turning point for air traffic management, and it sends a strong signal that the US wants to leverage advanced air mobility in the near future. But first, what is a m? Can you give us a brief explanation?

[00:38:39] Elena Zheng: So a a m refers to a new transportation ecosystem of small low altitude aircraft events, air traffic management and supporting infrastructure that aims to move people and cargo around more efficiently.

[00:38:53] Elena Zheng: It includes electric vertical takeoff and landing, also known as EV toe aircraft efficient propulsion [00:39:00] systems and technologies that enable safe, scalable, low altitude operations. Industry leaders are progressing towards initial a a m operations later this decade to name a few Archer aviations working with US cities on early trial flights under federal pilot program and is positioned as official air taxi partner for the 2028 Los Angeles Olympics.

[00:39:23] Elena Zheng: Joby Aviation has already delivered test aircraft and is preparing for early commercial air taxi services with some international activity expected as soon as 2026. Beta technology has completed fuel efficient passenger flight tests in New York with potential commercial services starting around 2026 to 2027, and W Arrow is developing what will be the first FAA certified autonomous passenger aircraft.

[00:39:51] Phil Doyon: All right, so Atlanta, you mentioned a few development, but are they already taking place in the US?

[00:39:56] Elena Zheng: So actually in the US most ETO aircraft are [00:40:00] still in the flight testing and certification phase, and that's because FAA certification process is very rigorous. Full commercial services will take time. As a result, several companies are advancing early developments abroad where regulatory frameworks and infrastructures are advancing faster.

[00:40:18] Elena Zheng: For example, Joby Aviation has delivered aircraft to Dubai and Archer Aviation is testing its ev to aircraft called Midnight in Saudi Arabia. Other companies are also conducting test flights and demonstrations to build early potential experience outside of us while continuing to pursue FAA certification.

[00:40:39] Elena Zheng: So back to the US roadmap, how did a a m roadmap come to be?

[00:40:44] Phil Doyon: All right. So this roadmap is the outcome of nearly three years of work, uh, following a congressional mandate that was, uh, allocated to the Department of Transport to convene an interagency working group to define what they call a national strategy for the [00:41:00] coming decade.

[00:41:00] Phil Doyon: So the working group brought together over 100 experts from 25 federal agencies, and the roadmap represent their consulate view of this effort. Now, the a a m roadmap actually consists of two documents. There is a strategy and a plan. The strategy says the larger policy visions of the United States, and it formulates 40 recommendations.

[00:41:28] Phil Doyon: The plan details how federal agencies will execute those recommendations over time. It's essentially a list of action items related to each recommendation and the agency responsible for executing them. Both documents are well worth reading for. Anyone interested in anticipating how the National Aerospace System, uh, may evolve over the next several years.

[00:41:52] Elena Zheng: So in your opinion, from human factors perspective, what should we pay attention to?

[00:41:58] Phil Doyon: So, what really stood out to me [00:42:00] reading the roadmap is how it frames the problem space, the strategy organized a m integration around six pillars, aerospace, infrastructure, security, community planning, workforce and automation.

[00:42:17] Phil Doyon: We can start with aerospace because that's a major one. So currently aerospace is managed by air traffic controllers that authorize what each aircraft can do. So a TC authorize you to climb descent, or, uh, move laterally. The roadmaps suggest that aerospace management will evolve toward what they call co cooperative areas.

[00:42:37] Phil Doyon: The idea is that third party service providers would manage defined portion of the airspace. That would still happen under the FAA rules and oversight, but it will no longer rely entirely on ATCs. So that's a big change.

[00:42:53] Elena Zheng: Yeah. So from my perspective, the infrastructure pillar, something like a call for action to modernize how we fly [00:43:00] the roadmap, ex explicitly calls for new surveillance solutions for low altitude, high density operations, and for communication protocols that move beyond traditional voice radio with a number of vehicles expected to fly, airspace needs to move toward data link communication to manage authorization and clearance more effectively because those can be handled by software.

[00:43:24] Elena Zheng: There's also strong emphasis on adapting existing infrastructure and regulations to accommodate new a m entrants rather than forcing them into a framework that was designed decades ago. The roadmap points out that the current airport guidance was never written with remotely piloted, supervised, or autonomous aircraft.

[00:43:42] Elena Zheng: In mind. Defining new facility and equipment requirements will be very critical here.

[00:43:48] Phil Doyon: Yeah. Another big challenge will be workforce as we are already facing pilot shortage in commercial aviation, and a a m will be competing for the same [00:44:00] pool of workers and we'll also create new job categories that don't exist yet.

[00:44:05] Elena Zheng: Right previous UAS workforce, um, studies identified dozens of new jobs across design, operations and support roles and schools and universities need to develop new education and training programs to prepare for those jobs.

[00:44:20] Phil Doyon: And finally, there is automation. The roadmap acknowledged the arrival of highly automated vehicles in the airspace like Whisk that Elena mentioned before, that is autonomous and remotely monitored.

[00:44:34] Phil Doyon: But we still need more information to know how to design and how to certify such highly automated aircraft. So the roadmap tasks, the FAA and NASA to research on human machine interaction and role allocation in highly automated and autonomous aviation systems, including performance during phenomenal and degraded condition.

[00:44:57] Phil Doyon: So we know that currently NASA and the FAA are [00:45:00] already working with the industry and academics via what they call the RAM AO working group. And so that stands for routine autonomous multi aircraft operations. And I'm really excited to learn about their results.

[00:45:13] Elena Zheng: Yeah, the AM roadmap has been well received by the industry with news reporting saying the major EV to players applauded the US government for taking a strong position on a a m in the country.

[00:45:25] Elena Zheng: And we're very excited to see how it will pan out in the future. So that's all we have for this week. Thank you for listening back to you, Nick.

[00:45:33] Nick Roome: Wow. That was great. I thank you, Phil. Thank you, Elena. That's, that's awesome. I love this. This is what a great segment.

[00:45:40] Barry Kirby: Yeah. We're clearly out of the job. We're done.

[00:45:42] Nick Roome: Yeah. Uh, you, you can, you, you can start next week. No, thank you. This is amazing. Uh, thank you so much to our friends over at the Aerospace Systems technical group for putting that together.

[00:45:52] Nick Roome: Elena Phil, thank you so much for breaking down the latest in aerospace systems. If you like that segment and want more you [00:46:00] can reach out to us. You could reach out to the, the technical group and let them know that you saw it here and that you love it, and, uh, give them some positive feedback.

[00:46:08] Nick Roome: All right. Let's get into, uh, the last part of the show. I guess that needs no introduction, but we'll just say it's one more thing. Barry, what is your one more thing this week?

[00:46:18] Barry Kirby: Well, it's not my, one more thing, it's actually my wife. Amanda's one more thing where she's this week got her certificate for her Master's of Science or MSE in Human Factors Engineering.

[00:46:30] Barry Kirby: So I'm just now pleased that actually one of us in the company knows what they're doing. No, I'm very proud of what she's achieved. She's not only got her MSC, but she's done that whilst we've had uh, obviously our children growing up and also running a and doing the day job at the same time.

[00:46:45] Barry Kirby: So to do all of that together. And, and she also reflected that on every assignment that she had, a child was either Ill or u Uni, issues at university or something like that. So there was never an element where she had a, uh, an assignment that [00:47:00] there wasn't some sort of family drama going on at the same time.

[00:47:02] Barry Kirby: Yes, no, and then she, her final project, she she did amazingly well on. So, no, I'm, I'm really proud. So congratulations, Amanda.

[00:47:10] Nick Roome: Congratulations Amanda. That's, that's huge. And, um, I feel like Amanda's part of the extended Human Factors cast family. Even though I don't talk with her regularly, I still feel connected to her.

[00:47:22] Nick Roome: And so I'm, I'm very proud too.

[00:47:25] Barry Kirby: I'll let you know

[00:47:25] Nick Roome: if it's worth anything.

[00:47:27] Barry Kirby: So what about you, Nick? What's your One more thing?

[00:47:29] Nick Roome: Oh my gosh. Okay. So I've had to keep this one under wraps for quite some time because if I had. Discussed it, then it would indicate that I was playing it. And, okay. Claire obscure Expedition 33 is a great game. I'm gonna start off with that. It's phenomenal.

[00:47:45] Nick Roome: It's, the story has stuck with me in ways that I can't describe unless you've played the thing. It's, um, very cerebral in a lot of ways. There is so much to love about that game. And, uh, the reason I couldn't [00:48:00] talk about it is because it was the first episode that we were going to do for user experience points.

[00:48:04] Nick Roome: And it was, it was specifically on the Perry system. And I'm just gonna play this for you. So if, if you want to go and find us on any of the social media platforms that have short form content, I will give you a taste of what to expect for our series User experience points

[00:48:23] Rydia Weiland: ready to pack, carrying an expedition 33 isn't about staring at the enemy like you're in an awkward conversation.

[00:48:32] Rydia Weiland: The Perry timing is locked to auditory hits using signal detection theory. This requires that players filter noise in order to detect signal. Your brain is constantly separating signal from noise. Sound cues cut through all the visual noise like motion. As a result, the game is designed to lower the and amplifies the signals, allowing users to filter through stimuli to detect signals, also known as auditory cues.

[00:48:58] Rydia Weiland: Perry. Timing is calibrated to [00:49:00] auditory reaction time, not visual latency. That's why it feels demanding compared to other games. Games like Dark Souls or Final Fantasy seven remake emphasize visual anticipation and motion cues. Expedition 33 leans Auditory, which the designers leverage to shape the learning curve for players.

[00:49:17] Rydia Weiland: This explains why dodging feels safer compared to Perry. The game gives you more grace, but when you man that Perry, man, it feels rewarding.

[00:49:27] Nick Roome: Yeah, there you go. That's a, that's a nice taste. Although the, the visuals on that one were cropped quite weirdly, uh, here on the live stream. So apologies for that. But you can go see the full thing on our social media platforms. Like I said it's a great game and, uh, the lab did a great job with that.

[00:49:43] Nick Roome: I'm incredibly proud of them. There's just a lot of pride here in the, uh, the one more thing this week. So congratulations, Amanda. Congratulations the lab, and congratulations to you all for making it through the episode. That's it for today, everyone. If you like this episode, enjoy some of the discussion about [00:50:00] AI and what it means for us as a society.

[00:50:02] Nick Roome: I'll encourage you to go listen to, uh, one of our more recent episodes episode 310, where we talk about how AI is reshaping human factors and UX roles, and how the seven emerging jobs are are coming out due to AI and decision making and all that stuff. So go listen to that, uh, comment wherever you're listening with what you think of the story this week.

[00:50:23] Nick Roome: Do you love the new segment, uh, as much as we do or more? Let us know. For more in depth discussion, you can always join us on our Discord community. You can visit our official website, sign up for our newsletter, stay up to date with all the latest human factors news. If you like what you hear, you wanna support the show, there's a couple things that you can do.

[00:50:40] Nick Roome: One, you can leave us a five star review, you can do that right now. Actually, uh, there's still a couple minutes left here this hour. Why don't you do it? Two, you could tell your friends about us. That helps us grow through word of mouth. And, uh, you know, people tend to take your recommendations a little bit more seriously than an algorithm or three, if you have the financial means to.

[00:50:59] Nick Roome: Why don't you [00:51:00] consider supporting us on Patreon? You can help us support our lab and do wonderful things like what you just saw and heard for user experience points and amazing other things that we are working on too that I can't quite announce just yet. So there's a teaser for you. As always, links to all of our socials and our website are in the description of this episode.

[00:51:19] Nick Roome: I wanna thank Mr. Barry Kirby for being on the show today. Where can our listeners find you if they wanna talk to you about, uh, ai?

[00:51:28] Barry Kirby: If you wanna come and chat to me about ai, come and find me on LinkedIn or Facebook, or if you wanna come and listen to me chat and interview people around AI and other human factors topics.

[00:51:37] Barry Kirby: Find me on 12 or two. The Human Factors Podcast at 1202 pod.com.

[00:51:43] Nick Roome: Excellent. As for me, I've been your host, Nick Rome. You can find me on our discord and across social media at Nick Rome. If you're watching live, stay tuned for the post show. For the rest of you, thanks for tuning into Human Factors Cast.

[00:51:55] Nick Roome: Until next time. It [00:52:00] depends.