Want this content delivered right to your inbox?
A conversation with Darren Cambridge, Managing Director of the Institute for Trustworthy AI in Law & Society (TRAILS) at the University of Maryland
“Friction is not always a bad thing. Friction is what keeps the car on the road and not going into the ditch.” – Darren Cambridge
In this episode of the New Friction series from the Facilitation Lab podcast, host Douglas Ferguson interviews Darren Cambridge, Managing Director of the Institute for Trustworthy AI in Law & Society (TRAILS) at the University of Maryland. They start from Darren’s framing that the next decade’s central question is capabilities and authentic humans together, not humans fitted in around AI. Darren distinguishes agency (having a voice) from autonomy (freedom within boundaries), and explains how explicit guardrails can support autonomy while still leaving people a say in where the lines are drawn. He draws on aviation, basketball, and football to argue that safety practices and rules evolve through iteration, and he warns that participation theater, where organizations listen but don’t share power, does more harm than good. The two discuss productive versus unproductive friction, how efficiency-first AI strategies can erode skill development and produce too much low-value output, and why organizations with strong participation and deliberation habits will handle AI more steadily. Darren closes with an invitation for practitioners to partner with researchers at TRAILS.
This episode is part of the Facilitation Lab Podcast. See all episodes
Show Highlights
[00:00:00] Capabilities Plus Authentic Humans Together
[00:03:00] Vertical and Horizontal AI Adoption
[00:06:00] Agency Versus Autonomy and Guardrails
[00:09:00] Rules Evolve: Sports and Aviation Lessons
[00:14:00] Who Is Accountable When AI Fails
[00:20:00] The Filing Cabinet Problem in Research
[00:26:00] Participation Theater Erodes Trust
[00:31:00] Productive Friction and the Cost of Efficiency
[00:38:00] Participation When the Baseline Is Unstable
Links | Resources
About the Guest
Darren Cambridge, Ph.D., is Managing Director of the Institute for Trustworthy AI in Law & Society (TRAILS), a $20 million NSF- and NIST-funded institute based at the University of Maryland. He is a cross-sector leader across higher education, nonprofits, government, and industry, and founded Mia Learning, an edtech AI startup. He previously directed the Connected Educators initiative for the US Department of Education, the Washington office of the National Council of Teachers of English, and content strategy at the American Council on Education. He is the author of Eportfolios for Lifelong Learning and Assessment, and his honors include the Booz Allen Most Promising Social Innovation Prize, the Wenger-Trayner Social Learning Leadership Fellowship, and the MacArthur Foundation Digital Media and Learning Prize.
Transcript
Douglas Ferguson: Welcome to New Friction, I’m Douglas Ferguson. AI just made execution almost free, so why are organizations still stuck? Because the friction didn’t disappear, it moved and it multiplied. It’s no longer in building, it’s in deciding what to build, how to align and how to move forward when the path isn’t clear. That friction, the human side of change, is what this series is about. Each episode, I sit down with leaders who are living it, navigating the real challenges of AI transformation, not the tools, the people. The task that took two weeks now takes two minutes. The work isn’t the bottleneck anymore, the conversation before the work is. That’s the work this show is about. Today, I’m with Darren Cambridge at the University of Maryland, where he manages the Institute of Trustworthy AI in Law and Society. He’s also the author of Eportfolios for Lifelong Learning and Assessment. Welcome to the show, Darren.
Darren Cambridge: Hi, glad to be with you.
Douglas Ferguson: Yeah, it’s great to have you. Looking forward to chatting further. It was great to meet you in DC when we had the executive dinner and excited to talk again. At the dinner, you mentioned a fundamental question for the next decade is capabilities plus authentic humans together. And a lot of AI rollouts, especially in the enterprise, posit the other way around. It’s like capabilities and humans fit in where they can. So why does that first framing demand of a company that the second one kind of leaves out?
Darren Cambridge: Well, I think one thing is that it frames people as not simply means to ends, so ends in themselves. So part of what any kind of organization produces is people, capabilities, experiences, those people sometimes stay with that organization, sometimes they move out into the other world, but we’re shaping the people who are doing the work in our fields, in our organizations and beyond. And there’s a fundamental moral imperative to treat those people as having value beyond just what they can contribute to a particular commercial or pragmatic. So, I think that it’s important from that perspective, from an ethical perspective. I think it’s also important because we frankly know more about people than we know about these technologies. The technologies are changing more quickly than the people are. There are some underlying realities about how people work together, what they value, how you organize them in organizations that don’t fundamentally go out the window even with a very transformative technology like the generation of AI that we’re working with. Now, we need to capitalize on that knowledge and then continue to learn about works at the intersection with these new technologies and how we make sure that what we know about people and organizations informs were guiding the unfolding of the new technology.
Douglas Ferguson: I want to come back to the point you made around the tech is changing faster than the humans. I think that it varies by organization. Some organizations are keeping up faster than other organizations, some are more slow-moving. I would say some organizations or probably the majority of organizations are heterogeneous in the sense that you got some folks that are keeping up faster than others, and so you’ve got this divide and it’s uneven capability across the groups. And so, I’m kind of curious to hear your thoughts more on the impact organizations and how we should be posturing to step into this time where things are behaving in these ways.
Darren Cambridge: Yeah. Organizations are in and of themselves internally complex. They have different organizational entities within them that shape who interacts with what and interacts with which technologies in which ways, but they also have cultures and more informal networks. The AI technologies, one thing that is particularly interesting about them is they’re coming in from both the vertical and the horizontal perspective. So we have organizations trying to introduce AI in a structured way to meet organizational objectives. And then, we have within some of these smaller groups, communities, individuals working independently, they’re also experimenting with these tools which are distinctively accessible to individuals out of the more formal structures for implementing and regulating new technologies compared to a lot of other similar disruptive technologies early on. I think about that kind of experimentation and at the individual level, the sense of some degree of agency that gives people that may be, in many cases, quite contrasted from the amount of agency they feel in shaping how the official organizational implementation works. There are really two change processes happening at once. I, as an individual worker, am thinking about using and collaborating with these tools to the sense that I think of them as having their own agency and then how that intersects with what from an organizational leader’s perspective might be being asked of me might be being facilitated by the resources and the rules that I’m asked to operate within.
Douglas Ferguson: Mm-hmm. Yeah, there’s lots to unpack there. The word agency itself comes up quite often, and I think there’s many even layers there around what kind of agency we’re talking about, who has the agency, what are the qualities of it? And also, that last piece that you mentioned around … well, I specifically, my brain went to governance and guardrails, which is something we’re thinking a lot about lately and how oftentimes people hear that word and recoil, but it can actually be quite liberating. It’s sort of like having a coloring book, so we know where the lines are and how to properly play the game, where is it safe to operate? It can be quite liberating to know those things because oftentimes folks given any options to do anything, it can be overwhelming and they don’t know where to start. And so, in a way, the constraints create a little bit of agency because they know where they can operate, so it’s like execute your agency here. And so, I’d be curious to hear your thoughts there just on all the various layers of agency and how it shows up in varying ways.
Darren Cambridge: Yeah. One distinction to think about an agency is my feeling of agency. Can I shape things? Do I have a voice? Do I have self-efficacy in terms of being able to have my efforts result in changes that I want to see in my work, in my organization, in the larger society versus autonomy? Do I have the right to work within my own domain and be left alone? Some of that sense of having structure and guardrails of what can I do, what lines do I need to color within can be helpful for me to understand my degree of autonomy. And often, I want autonomy within certain boundaries, and the fact that those boundaries are explicit and well-defined gives me confidence or increased confidence that I’m going to be able to act with autonomy within that space and that it’s not going to get pulled out from under me immediately. That doesn’t necessarily mean that I feel like I have much of a voice in shaping where those lines are set. When my experience and my professional judgment suggests that they’re in the wrong place, do I have means to influence improving the shape of that container in a way that allows me to do my work well and that is consistent with my values while still advancing the mission of whatever organizational context working.
Douglas Ferguson: Yeah, it kind of comes back to continuous improvement and what are the mechanisms where individuals have a voice in that continuous improvement process. It makes me think of how games would be really unfun to play if we didn’t have the rules. We just made them up on the fly being fun. But also if the rules are insufficient or inappropriate, they can also make the game less fun to play. For instance, baseball, if the bases were really, really close together, that wouldn’t be any fun. If they were way too far apart, it wouldn’t be any fun. But what if that was the case early on and the players didn’t have a way of sharing, communicating that, or there was no way of discovering and surfacing that up and then everyone has to live through it and the game probably has a short lifespan and people are miserable while it does still live.
Darren Cambridge: Yeah, if it doesn’t adapt. When basketball first started, there was no dribbling. You caught the ball and you either passed it or shot it. That was a limitation that needed to be addressed. American football didn’t have a forward pass as part of the game for several decades in its early evolution. I think no matter how many smart people like us get together and have these conversations and work on these issues, we can expect that it will be a process probably taking way longer than we’d like to sort this stuff out. People often in AI safety conversations talk about analogies to other industries like aerospace, for example, that have these well-developed privacy practices and safety practices and evaluation and approaches and critical incidents evaluations. That didn’t come along with the Wright brothers. It took decades of experimentation, some of it with pretty disastrous consequences for some of the people involved to zero in on what really works. And so I think we can spend a lot of time planning and analyzing in great detail, but some of this is a iterative experimentation process, which is why a lot of the processes that we have seen work well in other contexts or facilitating conversations and social learning processes, collaborative problem solving and knowledge generation in conversation with more formal kinds of academic research or measurement continue to be really important. They’re probably increasingly important in this kind of context because it’s out of those kinds of structured but not rigid conversations that we surface people’s actual experience in the context of practice with these tools and context matters. One of our fundamental hypotheses at TRAILS is that trust and trustworthiness are not conditioned solely of a system in isolation and they need to be considered in the context in which things are deployed and managed and governed and actually used. And those specific socio-technical contexts can vary quite a bit and those variations matter to what is appropriate trust and what you do to increase trustworthiness and also what you do to help people feel a sense of agency and actually have real influence on how things evolve.
Douglas Ferguson: The aviation example really struck me because a lot of the techniques now were informed by early black box recordings. It’s like there were causes of crashes, oftentimes not just mechanical but collaborative human issues that weren’t discovered until they actually listened back to black box recordings of what was transpiring prior to the crash. And so, telemetry is important and we’ll certainly learn from that in these AI systems, but hopefully, we don’t get to these catastrophic failures to learn those lessons in this case.
Darren Cambridge: Yeah. Black box are important, but also going back and looking at what were the PowerPoint slides about the O-rings, the larger context of organizational or sometimes multi-organizational conversations that inform what you see in the interaction between the technology and people that you’ve measured directly in the AI equivalent of black box, which might be different ways of trying to create an audit trail. The email messages between people that were involved that don’t directly intersect what’s happening with that system may also be really consequential to understanding what happened, so how do we create an environment where we have enough of that context in action documented that we can go back and have a fuller picture to understand of additional guardrails or interventions would’ve been helpful to keep things on the right track?
Douglas Ferguson: Mm-hmm, yep. You mentioned earlier horizontal versus vertical, and I wanted to dig in a little bit more on how you’re seeing that. Those terms are used for a lot of different things. I’d love to get a little more understanding there.
Darren Cambridge: Yeah. Well, I’m thinking particularly of the version of that that is about authority and accountability, that I have vertical accountability to my boss, to my boss’s boss, to the formal governance structures in which I’m working and the people in positions of authority. But then, I also have horizontal accountability to my peers, to other people that I’m interacting with more informally, more directly day-to-day. Part of that is tied up in many cases in shared professional identities, and what does it mean to be a lawyer who works in a highly AI-engaged kind of organization? There is a longstanding existing professional identity and category there that people are really deeply invested in, but there’s also a new emergent one within that kind of context that’s very much being defined, arguably is going to be defined as much or more so by those horizontal conversations as it is by the interaction with the vertical power structure.
Douglas Ferguson: Mm-hmm. That makes me think too about your comment around orgs being complex and the structure of the organization is often an artifact of how our human-shaped processes and roles existed in the past. Quite frankly, a lot of it was influenced by Taylorism and growing out of the factory work optimization. And then now, we’re introducing these AI tools that pretty powerfully integrate stuff across the value chain. Organizations have rarely been organized or structured around the value chain, so I’m kind of curious if the research, your work you’re thinking has focused on this concept of the AI being able to cross-cut across functions and follow the value chain from start to finish versus having to go through all these handoffs of various functions, departments, and the trade-offs about embracing an approach like that versus mirroring the existing structures.
Darren Cambridge: Yeah. I don’t know that we have a lot of research that can speak to that level of organizational transformation. Certainly, an intriguing phenomenon in the organizations where that is happening. I think in reality, right now, it’s fairly few places that really have created the context for that kind of deep transformation. And often, they’ve done it in ways that not always turned out to be the best. There’s the famous example of Orna laying off lots of folks and basically making the decision that we mostly don’t need people and doing an IPO and then realizing, “Oh, we kind of need some people.” How they plug in may be a little bit different, but the way that that played out did real damage to people’s lives, to relationships within that organization, at least in the short term, to its ability to efficiently execute its mission that some of that may be a necessary part of transformation is try stuff that doesn’t work. I think it is more likely to be effective if it is grounded in thoughtful participation of the people who are involved and consideration of the social and relational values that underlie the work.
Douglas Ferguson: Yeah. To me, the structure of an organization, it comes, I don’t know if you’ve heard of Conway’s law, but this idea that any software systems we build or really any systems we build are going to mirror the organizational structure of the business or enterprise or institution. And it’s only natural because we’re humans making things, and if we’re organizing, we’re communicating in certain ways, then naturally anything we build is going to follow those same boundaries and resemble those same structures. And if we build our AI in those ways, then there could be some real limitations and it’s their ability to see broader patterns or fix more systemic issues. And I agree with you that we shouldn’t use its capability to be cross-cutting and value chain innovating to say that we just replace the humans, because there’s plenty of need to have humans. One thing I’ll get your thoughts on is something we’ve seen lately coming up pretty frequently is this idea that, this conversation around what should the humans still do or what’s the human’s role in this work going forward? One sentiment that comes up often is that AI can’t be accountable, which is an interesting idea here. But at the same time, you’re seeing legal cases where they’re saying that AI can be treated as a person. There’s not a lot of litigation per se on that, but there’s some case law out there already, if I understand correctly. And so just curious to get your thoughts or have you seen any things around this idea that want a key differentiator or are there others we should be thinking about?
Darren Cambridge: Yeah. You can certainly point to AI as being responsibility for some kind of failure or incident that’s problematic, that that is very much grounded, in some cases, how the technology is designed, more often it’s about how it’s deployed and what it’s allowed. But ultimately, accountability requires consequences that meaningful consequences that are similar to the kinds of consequences that humans and organizations face for a technology is a challenging thing. Turn it off. It doesn’t care because it doesn’t care. Care is a fundamentally human characteristic. It requires embodiment, it requires a history of social engagement that can be simulated, but not, at least anytime soon, replicated with technology. So ultimately, the person or group of people for which there needs to be consequences or a particular use of a technology to be accountable, how you figure out who that is gets a lot more complicated in Agentic human-AI collaboration context. That strikes me as a complex problem, not a fundamentally complex in the sense that we’re never going to be able to find a solution. These things are traceable. It is to some extent still mechanistic. We have a lot of work to do to be able to describe context, document context in a rigorous enough way to allow to trace those chains of accountability and identify where the buck needs to. So that’s a really important area where we need a lot more research.
Douglas Ferguson: I like that you referred to it as, I’ll quote, “It’s still mechanistic.” That kind of struck me because a lot of folks talk about it, deterministic versus non-deterministic. But at the end of the day, it is still … there’s a mechanism, we put something in and we got something out and that is traceable, knowable. I think that’s something that maybe people lose sight of when they get so focused on it being “non-deterministic.”
Darren Cambridge: Yeah. It’s traceable. Something was asked to do something and given a certain range of options, and then something happened and that intersected with something else. We can document that. We may not be able with absolute precision to say, in this chain from point B to point C, why did the thing that come out at point C versus what came in at point B? Some of that stuff is not deterministic, but it does not mean that fundamentally the larger pivoty system in which that technology is functioning is somehow unexplainable and interpretable. I don’t think that’s the case in many-
Douglas Ferguson: Yup. I want to come back to another thing that you said at dinner. I don’t know if you invoked the actual “filing cabinet” problem name, but I believe you did talk about researchers not sharing unvalidated things, and so this messy middle never emerges. And I think also, I’ve heard the filing cabinet problem described too, is if they do prove something, but prove something that they weren’t hoping to prove or they disproved a hypothesis, that rarely gets published either because that’s not how the publishing system works. There’s a vast amount of context and knowledge that is just inaccessible to other researchers. And then now, it’s also similarly inaccessible to these agents and knowledge systems.
Darren Cambridge: Yeah. I think we’ve known for a long time there are different [inaudible 00:25:53] to in forms of knowledge, and certainly, formalized research is really important where we can get it. And, stuff that disproves hypotheses may be as important as stuff that doesn’t. One of the things that we’re struggling a little bit with is that we have done a bunch of research on participatory processes in governance and design, and a lot of that has said a lot of the things that we assume will work don’t. And that stuff ends up being more, participatory processes can be more like decision-making because there isn’t a strong commitment to actually listen to and act on what’s being said by the people who are invited into these participatory processes. We have one study, for example, that looked at various participatory mechanisms, federal governments around the globe around AI, and in very few cases was there actually any evidence that what happened in those processes influenced what policies were set, what guidance was issued. So we don’t really take those processes seriously. We know what works and what doesn’t work. That’s not something that, as researchers, alone we can create a context where we can evaluate that. There has to be organizations, governments that really make that deeper kind of commitment or we can determine the specifics of how things ought to work well in terms of creating those opportunities for agents.
Douglas Ferguson: Yeah, I’m curious, was the finding just that those experiences didn’t lead to the outcomes suspected or that the contributions didn’t find their way into the policies, or was there a layer deeper where there’s some understanding of why it failed or some of the causes of the unexpected behaviors?
Darren Cambridge: Yeah. I think, in large part, it was because genuine participation requires sharing power in a way that in many of these organizational and governmental contexts, there wasn’t really the clear commitment to giving away some power or to distributing it more broadly. Even when maybe the rhetoric was that that’s what you were doing, there may have been good intentions, but when the rubber hit the road, you could not see where that connection had been made. A big part of it is, if you’re going to do a participatory process as an organization, you really have to be really clear about what that means in terms of you’re flattening your decision-making in a way that would be challenging if you don’t like what comes out of the deliberative context in which that you’re creating. My sense is it does more harm than good to have these kind of participation theater.
Douglas Ferguson: Yeah, it certainly erodes trust, because if we say we’re going to engage in this process and then it’s clear that it’s more talk than it is action, and then people begin to not trust that process, they begin to not trust anything we’re doing around the project itself.
Darren Cambridge: Yeah. This is not nothing new with AI. I think every listener probably can think of a context where they work with an organization and there were lots of listening sessions and surveys, and “We want your input and ideas on X,” and then senior management did what they wanted to do anyway. There has been a history of those kinds of processes not being trustworthy, which we’re seeing mirrored in what’s emerging with AI that has to be addressed at the significant organization.
Douglas Ferguson: Yeah, it’s interesting. We talk about the whole premise of the show is new friction, but something we talked about early on as we were picking the name and we said a lot of this stuff is friction or organizational dysfunction, things that we’ve been dealing with for years and either people ignored it or maybe it wasn’t being experienced at a level or a volume that was detrimental. You could wrap some process tape around and ignore it, but AI is certainly amplifying a lot of these things. It’s interesting to hear you talk about where we’re seeing these processes fail. They’re failing in very similar ways before, it’s just in this new context.
Darren Cambridge: Yeah. I think friction is not always a bad thing. Friction is what keeps the car on the road and not going into the ditch. It’s the right friction on the right [inaudible 00:31:22]. Having genuine participatory processes, that is not always efficient. It takes resources, it takes time. You have to build relationships, you have to take all of this data from complex conversations and figure out what it means and what you do with it. But I would argue that that kind of friction can be very productive versus if you do the theater route or just skip that altogether, you’re getting less productive friction because it’s based in distrust of the process, distrust of the organization, feeling like people are being used as a means to an end, that their expertise is not used as if it were currency, but also being appropriated and incorporated into technology in which not only they don’t have a voice, but it devalues the expertise they have as individual professionals.
Douglas Ferguson: I love that you went to this some friction is helpful. We have a concept of friction discernment, and so how do we sharpen our ability to notice what things we want to keep in the system that might be inefficient? I think that actually echoes something that you brought up at dinner in DC, which was around this notion that creativity and innovation demand inefficiency. Right now, the dominant AI business case is efficiency. In fact, I just heard about it. I was at a conference and a company was talking about they put the head of AI sits underneath the CFO because they view AI’s purpose is to improve efficiency. And so, the CFO is efficiency-minded, so let’s put him there. I’m curious, if you have any stories or just any perspectives on what an organization loses when it optimizes around efficiency to the detriment of other things, how long does it take for them to notice?
Darren Cambridge: A lot of my background is in education and learning, and so my mind goes to the emerging body of research around AI at the intersection with teaching and learning and professional development. It can make a lot of things easy in ways that don’t help people develop their own capabilities and the role that they can play distinctively contribute to the larger system. People write better with AI when they’re already good writers, and that means sometimes you make strategic decisions about when to use it and when not to use it and make certain problems harder for yourself. That may be a less efficient writing process that’s one helps you develop your own thinking more effectively, helps you develop your ability to discern the quality of what you get. You’re working in collaboration with the systems in a way that if you went the most efficient route of saying, let’s let the technology do everything that it can do reasonably competently, you lose that and you lose a sense of investment. And also, frankly, when there’s little friction on producing stuff, you may produce too much stuff or the wrong stuff easily because you have less skin in the game. I can produce all sorts of reports that might be helpful to people across our institute. We have a real challenge of 80-some faculty members at four institutions in almost 20 disciplines that are doing both disciplinary and interdisciplinary work, and we’re trying to weave it all together and get those folks connected to each other. I have tons of documentation of that. I can use AI tools to present lots of different views on that, but having more of those documents, analyses, decks, whatever is not necessarily better. It can sometimes be a distraction from figuring out how does this system work and what are the right places to intervene and where is less more in terms of creating a context where people are going to interact in a way that gives me a deeper perspective on what’s going on that I could not get from the stuff that is cured and mediated and in a form that you can process for me.
Douglas Ferguson: Another thing I was thinking about was something that came up at dinner around, I think the word Molotov cocktails was even mentioned around leadership throwing these bombs are Molotov cocktails around in every direction and people don’t know what’s safe even without AI. When the baseline is already that unstable, what’s happening when we layer an AI rollout on top of that? Really curious if there’s a version of participation that works at that point, like the foundational layer that is already rocky or unclear or unstable, or does the participation work have to come in later?
Darren Cambridge: I would argue that creating that social infrastructure for collaboration and learning and deliberation around what counts, what matters, what should we do? Do we do that in ways that takes advantage of the capabilities we have individually and collectively and there’s the people involved in the shared and individual values, that work is at least as important and is not specific to AI. I don’t have research that I can point to support this, but it would be my personal hypothesis that organizations that develop those kinds of ultras of participation and deliberation and shared power and distributed knowledge development are going to have an easier time thinking about how to appropriately incorporate these new technologies. And there much less likelihood in those kinds of environment that there’s going to be the Molotov cocktail throwing either down from the CEO down or increasingly we see that at the horizontal level too, people resisting in ways that can be problematic both in terms of organizational outcomes, but also in terms of the culture, the organization that you would like people to not feel like they need to go to the barricades because there’s somewhere else to go.
Douglas Ferguson: This is making me think about a couple of things that came up earlier in the conversation. If organizations do engage in the participatory work early on, or maybe even just revisiting, if they have a program that’s already going and not working well, actually starting with the governance and inviting participatory work there hits on two things we talked about. One is the reinventing the game rules or evolving the rules based on what we know and getting comfortable with that. But then that also has to be paired with this ability to truly let go of power and let the process drive the direction and rules and guardrails and things that we put in place. And if they’re willing to do those two things, I agree with your hypothesis. It would likely lead to better outcomes, better governance, and a more well-aligned team, an org that can move more fluidly and have more confidence as they’re doing this work.
Darren Cambridge: Yeah, absolutely. Those are not new challenges. Those are fundamental organizational leadership challenges that we’ve been talking about for several decades and important leadership perspectives on how you guide and support innovation in complex internally and externally connected organizations continue to be as relevant as new technological.
Douglas Ferguson: Darren, I could keep going for a long time, but we do have a time constraint here, so we’re going to have to come to a close. I want to give you an opportunity to leave our listeners with a final thought.
Darren Cambridge: I guess I’ll say, working now in academia with researchers, researchers really want to help shape how some of these challenges unfold. They don’t always have great tools and skills to make connections to people in industry, in government, in other kinds of sectors. But the vast majority of the time, there is a real interest in being helpful, connecting, particularly when there’s an openness to multiple perspectives. So please, find your local researchers or national resources. Reach out to us at TRAILS. We’d love to talk to you and collaborate with you. We’re actively trying to figure out how do we take what we have learned and are learning about trust and trustworthiness in AI across multiple contexts and share that in a way that’s helpful for people within organizations and within governments who are making consequential decisions about how AI further gets developed, how it’s deployed and used and how it’s governed. So talk to us, tell us what you need, what would be helpful, and we can figure out together how we take the body of academic literature, which is large and growing on a lot of these topics and put it to work in the organizational context in which you work, where you have much richer understanding of how it works than almost any researcher can being outside of those contexts.
Douglas Ferguson: Great call to action. I love that, Darren, and I hope the readers do take you up on it. I think we need to see more public-private collaboration. And to your point, researchers need access to great cases and data and stories and maybe even ride-alongs and opportunities observed. So please, do take his call to action and embrace it and reach out and get involved. Darren, it was great to have you on the show and really enjoyed the conversation.
Darren Cambridge: Yeah, thank you very much, Douglas. This was great and look forward to continuing to participate in conversations to come.
Douglas Ferguson: Thanks for listening to New Friction. If you enjoyed this episode, share it with a leader who’s in the middle of this right now. They’ll thank you for it. And if you want to go deeper, we bring leaders together through executive dinners and virtual masterminds. To learn more about our work or to inquire about exclusive executive events, visit voltagecontrol.com. I’m Douglas Ferguson. See you next time.
