HRExaminer Radio Executive Conversations Badge Podcast Logo

HRx Radio – Executive Conversations: On Friday mornings, John Sumser interviews key executives from around the industry. The conversation covers what makes the executive tick and what makes their company great.

HRx Radio – Executive Conversations

Guest: Kyle Jackson, co-founder and CEO, Talespin
Episode: 341
Air Date: September 27, 2019




Important: Our transcripts at HRExaminer are AI-powered (and fairly accurate) but there are still instances where the robots get confused and make errors. Please expect some inaccuracies as you read through the text of this conversation. Thank you for your understanding.

Full Transcript with timecode

Good morning, and welcome to HRExaminer’s Executive Conversations. I’m your host John Sumser and today we’re going to be talking with Kyle Jackson who is the co-founder and CEO of Talespin an Enterprise solution software company that does virtual reality. You may have seen tailspin in the news lately.
[00:00:32] They’ve got a virtual human being who you can fire over and over and over again as you learn to do difficult conversations. How are you Kyle? Good job. Thanks for having me. Yeah, it must be crazy to be the inventor of a guy you get the fire horn over again. It’s becoming interesting that it’s become an interesting demo for sure.
[00:00:55] You know, we put that out there as a bit of a task this to see if you could actually have an emotional experience to with a virtual human and obviously one to pick a topic that was familiar and uncomfortable and and and you know the same time Universal and turns. That yeah, it can be really highly effective and ultimately, you know, we’re not building simulations to fire people really we’re not maybe that’s not even being sold.
[00:01:20] It opened the door for things that are more around, you know, the difficult conversations that actually should be empowering, you know, teaching people manage their feedback and teaching other leadership skills and interviewing skills and negotiation skills and all the things that you just you know, you do need to practice at but there is a there is a strong component of nonverbal communication and emotional realism that.
[00:01:42] He makes you good or or you know mediocre at those kinds of skills. Well up, you know, it’s it opens the door to the interesting kind of learning that I don’t think people have really thought about before, you know, the difference between a young manager and a seasoned manager often has to do with the ability to make sense out of people’s potential and how to develop it and you can’t see that you can’t see that until you’ve seen it.
[00:02:08] Five hundred times to say and and what you can see after you’ve seen it 500 times is pretty amazing and the possibility that you could simulate that kind of emotional engagement in a way that allows you to accelerate the learning of that kind of judgment. That’s a really interesting idea. Yeah, and we’re definitely not there yet.
[00:02:29] I mean that that is I mean it’s so exciting but it’s also you know, we’ve got a lot of new ones to build into the various scenarios tip to get to where you really are identifying those kind of those markers, but obviously even just even just understanding what are the best practices of giving. You know constructive feedback and having a place to practice that in the retires before you get in and you start to Bumble it as a manager and you black out and forget all your best packing all your best steps.
[00:02:56] It’s like at the oh how else would you build that muscle memory? So it’s that, you know, the kind of simulated nature of it and and also just kind of the the ability to get you know, exponential experience on a number of things not just conversational training before you actually get thrown into the line of fire is.
[00:03:15] I really different and new, you know new way to approach training and if we don’t really quite fully know how to that’s going to affect, you know, learning overall because you start to look at the types of exponential gains that we’re getting and it starts to challenge. You know, how the other modalities still play.
[00:03:34] So it’s really so yeah super interesting time. So let’s let’s start from the top. Why does take a moment and introduce yourself and and and tell us about how you got here? Yeah, cool. So I spent the most of my you know, my career prior to this been in meeting their attainment and I was really building a lot of kind of core infrastructure for emerging workflows Caster technology and then distribution technology.
[00:04:02] So, you know as we went from the analog. Do digital and you know, the early 2000s, you know, the entire media industry had to kind of be re-architected and I got to play a part in a bunch of different, you know pieces of that and I guess ultimately it kind of put me on a path towards being in the thick of it when this whole XR wave started up in 2012-2013.
[00:04:33] And so it started out as really. You know kind of from the technology side being one of the capable people who could figure out, you know solutions to really complex workflows and all the types of things we would have to overcome. And at the same time as I was as that was happening and it was just an exploration of the technology for Technology’s sake, you know, and the 2012 to kind of 2014 timeline.
[00:04:59] I was also spending a lot of time looking at how Ai and and computer vision and machine learning was going to affect work originally got into that because I was you know kind of interested in that as well. What would it mean? We could detect things, you know objects and things as it relates to meeting their attainment or other other were close.
[00:05:23] But soon as you get down that rabbit hole, you really start to understand the landscape of how it will affect productivity human productivity and then it’s just that just opened up BLM Pandora’s Box for for looking for ways to help. Offset some of that and I’m just so happened that this other technology that was that was I was working in in terms of ER and Mr.
[00:05:46] Game to be pretty powerful for what it could do and so started looking at you know, suggesting ways in which simulation and immersion and the types of tools that we’ve now been creating for years could help to, you know, transfer knowledge more quickly or prepare people more rapidly because. The Assumption was that was going to you know, our place and part of you was going to continue to evolve at an increasing pace.
[00:06:12] And therefore we defined a faster way to keep up. So it really landed us really, you know, kind of like, you know one was a really a real passion for for. The kind of macroeconomics vexed and problems that are out there as it says of retains this whole future of work conversation. And then the other was obviously, you know, 15 plus years of skills and experience that happened to be relevant with this new medium.
[00:06:39] That was so powerful and so down the rabbit hole we went and here we are, you know, four and a half years later having built some stuff that’s really getting exponential results. And it seems to be that there’s there’s there’s a there’s a real thing here and it’s kind of exciting. How it’s going to start to play out more.
[00:06:58] So what exactly does your company do? We’ve been you know, so we’ve been building, you know VR. And mixed reality primarily Business Solutions, you know to basically rapidly train or give people experience as in the simplest terms and so on the VR side that looked like stepping into a simulation like the conversational simulation that we talked about or maybe entering a customer’s home to solve the problem and being taught like investigative skills.
[00:07:31] Yeah, but I physically in present in that scenario and I play it out. Now again, my experience or scoffs kill myself and then on the other side and mixed reality. We’re going Solutions where we start to actually take a lot of those. What would consider B2B job AIDS in the past but really put them in the line of work.
[00:07:50] So when you are right in the heat of action, you know, and you’re trying to remember what you’re supposed to do around an installation or a certain process or again investigating something we can bring up assistive. Gee that helps you to be confident in those decisions. And so really it collapses the learning and execution stack is where we really are going to go and that that that starts to open up new ways to think about, you know, how frequently we can move and how comfortably we can move between positions as work continues to evolve.
[00:08:27] So how do you use a high or machine learning or whatever is in that bucket of near Magic that people talking about these days. How do you use that stuff to get you worked up? We use it a couple of different places and and it’s and it’s still, you know, relatively early in both of them you so you could imagine in the conversational simulations.
[00:08:52] We want you to use your voice and we want you to use your natural posture and your and be able to detect sentiment and have all of those things as inputs to affect the scenario wherein so there’s a lot of you know, natural language processing natural language understanding. We do have quite a bit to allow you to have a flexible conversation where we map what you said too intense.
[00:09:15] So it doesn’t it can as we can go let things go A bit off script so you don’t feel like you’re just in there doing point-and-click, you know lessons. So on the on the conversational side, there’s quite a bit going on there on the training simulations when it comes to process and we bucket things into a couple buckets.
[00:09:36] We start with object-based. In which is just kind of like learning the parts of things. What we found is pretty much every company when you step in there’s a lot of different nomenclature and Basics you really need to kind of quickly get up to speed on and that can be a blocker for getting into the more meaningful stuff.
[00:09:56] So when it comes to object based learning, there’s quite a bit you can do around object detection object segmentation using, you know, you computer vision and machine learning to help to accelerate the real some of that. You know, how we use that or how we build those into English in the process stuff, which is the second bucket process based learning a lot less.
[00:10:20] So, right it really gets really get specific into the process. Tend to build systems that are randomized evil and sin more scalable. So you can scale the amount of experience you can get but it’s not really in that case. It’s more like procedural systems than it is AI but this, you know, all of these Technologies, you know play a part in trying to get us to our end goal, which is really render a high quality experience that has the range of diversity that it really allows you to truly, you know.
[00:10:56] Achieve that new skill quickly versus what we see in the past, which is obviously very quick through or powerpoint-based or video-based One path type of learning and so because of that variety we start to open up new Pathways in the way, you know, we learn in a encode experience and so it does take quite a bit of machine learning and AI to go down that path.
[00:11:22] So you’re doing. What would I would think of is pretty Advanced research and development and that’s made easier by open source Technologies. How do you think about open source stuff? And and and and are you building on? Sort of existing public knowledge to get to where you going. Yeah. I mean you have to write as a small company.
[00:11:51] I mean, we there’s no way we can build a, you know, a natural language understanding stack ourselves on top of everything else. We’re trying to do when it comes to rendering, you know, an expressive virtual human. So so there’s a lot of there’s a lot of bringing Technologies together. We work from a from a for a lot of our stuff we.
[00:12:12] We build on top of an a game engine called unity and unity has a very active community of people who contribute back into a an asset store where people can buy and exchange each other’s kind of subparts. So there’s some of that gets shared, you know, openly and and because become some of the low lot low level stuff for both the our virtual human technology and and the.
[00:12:40] A job simulation technology when it comes to the actual speech and and AI stuff we’re you know, we’re very aware of kind of, you know, we’re building enterprise software and and ultimately people need to feel safe and how this stuff is being developed. And so we’ve been leveraging, you know, some of our significant amount of the zoo or stack but Microsoft for four.
[00:13:09] Both infrastructure and for how we get how we you know, how we leverage machine learning. So we’re not going as far as just completely open source stuff there that would add another layer of discomfort for companies because they don’t quite understand. He’s the who’s the board that’s behind it that is checking the stuff.
[00:13:30] There’s some really cool experiments out there and we get a lot of inspiration from them. But but we have to be pretty cautious about going down those those those routes because of. Yeah, this having been burned in the past by project sun setting and also into critical part of your software because where we are we are doing on kind of edge Rd by the other day.
[00:13:50] We do need it to get into customers hands very quickly. So I can’t just can’t just go in there and then and then Sunset, you know, it has to we have to be up really confident that it’s going to be around and it’s going to live on as a project or continue to advance over over, you know the coming years.
[00:14:09] The kind of work that you’re doing I am I think it I think it is a sort of a Bellwether thing that opens the doors to a lot of ideas. We haven’t really did jested yet. Right and so so the first the first question in that area is safety, and I wonder how I want how you think about safety with your work.
[00:14:40] Well Physical safety psychological safety, there’s there’s physical safety of course, but that’s kind of a machine interface question. I’m much more interested in sort of psychological and organizational safety. It’s really it’s a good question and it’s a tough one because. So I was talking with with the weekend.
[00:15:06] We just came from the Oculus convention this week, which is you know, the VR development communities convention and I was talking yesterday with the colleague there about about this a little bit and you know, we start we start to open up whether it’s there’s two different sides to it. I think there’s a side of what if we go right and so if I look at.
[00:15:31] Kind of the unknown on unconscious, you know discomfort. That’s right below the surface as it relates to where we’re of many of us are going to fit in the future of work and how our dogs are going to be affected by some of the technology and Innovation that’s happening. I think that’s a underlying kind of safe psychological safety issue already and the more I talked to companies there.
[00:15:55] There’s there seems to be a just an uneasiness about how to design where this is all going. What we’re trying to do obviously on the other side of that is provide some some consistent framework for for relative, right? And so, how can I quickly move between whatever it was I doing or was doing to what I need to be doing in the most empowered possible way now that in its first phase, you know, we’re seeing that reduced training times from you know, from weeks two days or four months two weeks.
[00:16:30] And that’s that’s pretty cool and its second phase. If we really get out get out there and we start to think about well what happens if I can learn skills that quickly it opened up a little bit of additional kind of gray area around. Well how much of how much of our work is going to move towards gigs versus careers because you know, either people might want to move more freely because we’ve seen that Trend but we’ve also seen the.
[00:16:59] You kind of end and and dark side of the gig economy to a little bit where people feel they don’t have community and they don’t and they don’t feel like they have a foundation to build our lives upon. So I think each one of these these topics has like really a there so there’s such a duality to them and and obviously that’s partly because there’s people who love both sides.
[00:17:22] People there’s people who love the freedom of the gig economy and there’s people who are scared to death, but they don’t know if they’re going to be able to make rent because they can’t consistently predict what we have the number of gigs are going to get and so the real weird time for the some of these Concepts and I think the stuff that we’re working on is generally just trying to make sure that people are are empowered with with you know, with with the skills that are in demand so that that problem or that.
[00:17:51] And maybe become something much more predictable. Well, so you’re going I think you’re going to end up hitting questions that we don’t know exist just yet. Right and like bike safety is such a gross way of thinking about the little nuances that you. Teach without knowing you know, this is this is the the most interesting thing to me about learning.
[00:18:26] Is that is that what gets taught and what’s in the curriculum are often not the same thing and difference between those two things is that people learn more from what the instructor or the instruction medium does then what it says. And so the capacity to transmit. Bad information down a training pipeline is is has always been pretty high but as we get into these hyper intimate hyper nuanced kinds of trainings like you’re talking about doing there’s this there’s this crazy thing about not knowing what we’re about to stumble into and it requires a kind of courage in the company to go down that rabbit hole.
[00:19:18] For sure for sure and it’s it’s something that we obviously think a lot about being very careful also about the topics that we invest in right? So, you know what we’re looking at we’ve been pushed towards things that are definitely I think much more sensitive in terms of building in you know, a bias or you know building in bad learning as into a more systematic are systemic.
[00:19:47] He had a way of doing things. So, you know, we’ve stayed away from some of the things when it comes to diversity and inclusion or when it comes to harassment because you know, I think that’s too nuanced and and is not it’s not about process. It’s really about you know, really starting to have meaningful conversations to change Behavior over long periods of time and it’s not like put somebody in simulation.
[00:20:17] Come up to you know to this new you but there’s definitely you know, there’s definitely a lot of stuff. That’s not that and so I think you definitely take a lot of care into the which topics, you know, we go down we go down this path in terms of the tools we have today. I get that stuff being you know impactful in the future.
[00:20:41] But but the technology is not ready for that even though it can be a you know, a good buzz word to tackle in terms of business. We’re being very cautious about where the stuff gets pointed because our our end goal is is really, you know, human empowerment, you know, not not too hard encode, you know, bad behaviors and so.
[00:21:06] I think once you start to really edit like it takes a more granular understanding of the technology stack and and where we’re at and what we can accomplish and what we can measure to be have a very clear lens on the yes, no around certain topics. But internally, we definitely have that and so, you know, we’re looking at things that are obviously much more kind of assistive or helpful, but not, you know, potentially as damaging when it comes to if it gets if we got.
[00:21:37] That’s just that’s just that’s a scary. It’s a scary proposition. Yeah, so so this is a hyper-competitive environment for development talent and you’ve got this sort of in addition to needing the kinds of people who are in the greatest demand. You’ve got kind of a. An ethical quality filter on top of that you don’t you’re not going to want people who solve problems just to solve them.
[00:22:07] You’re going to want people who show problems with some care. How in the world do you find them? And how do you compete for? Is it interesting? You know, we’re also bringing, you know, so a lot of what we use at a lowest level has been historically using meeting or timid or gain gain, right?
[00:22:22] Because the real-time engines were developed in those those field and so a lot of the people that go in to do that kind of work, you know, they their eyes haven’t necessarily been open to what we’re talking about. And so that’s definitely a challenge but there’s I think there’s a there’s a there’s a growing.
[00:22:43] What we’re finding is as consistently kind of a growing I wouldn’t call it like entity the people who are just entering their career with people who have been in their career for college, you know, five seven ten years. There’s there’s an Awakening there in terms of the power of the tools that they’ve been taught and that can that can be in the machine learning field or the the computer vision field or in game development.
[00:23:07] And we find that that’s where the talent really wakes up to want to work on something like what we’re trying to obviously we would love we would love everybody to to want to jump in that you know in the water and help us but. What we saying is it takes you know, because we’re pulling people out of what they what their mind they had their Mind Set On and into a new realm.
[00:23:33] Yeah. It’s it adds an extra layer of complexity in terms of getting development Talent. But you know, we’re finding pocket. We’re Global to by the way. So that’s the thing is is, you know at this point you’ll have to go Global for talent. So we’re already we’re already basically working in five offices to in Europe one in Canada to in the u.s.
[00:23:54] Always kind of their own specialization. So we’ve started to try to you know, try to go where the talent is, but it’s been it’s been a definitely I think it just like. Ready else, you know, I mean this is this is one of the key critical elements of anybody being successful right now is finding the skills and the talent to work on problems that they didn’t even fully understand that their skills are relevant for before.
[00:24:20] So we’ve spent a bunch of this conversation has already touched on the ethical issues in the work or other. Are there any key ethical issues you’d like to highlight another part of the work.
[00:24:34] Well, we you know, there’s some interesting stuff obviously around measurement and and I kind of like people analytics that obviously is is another big open area here and and I think you know again this is one of these these these touchy, you know could could go either way kind of topics, but you know, I think that getting to like a true.
[00:25:02] More a more accurate evaluation of somebody skills or competency when they’ve been given an opportunity to practice because the simulation gives them the chance to really, you know, get more proficient at that. That’s an interesting area. Right? So I look at the way or the discussions we have with large corporations and and how they measure Readiness and then and then how ineffective that measurement for Readiness can be because the terms of.
[00:25:31] You know months or weeks later are too high. Like you should have been able to predict that that person was either not a good fit or that was they were not ready. And so there’s it’s it is an ethical area because obviously the deeper you get into measuring anything the more, you know, it’s a little bit scary, but I I think the offset to that is is the empowerment part which is give people the opportunity to practice.
[00:25:59] Don’t just throw them to Something in. For them and then and then make you know ours decisions Empower them with the tools. They actually get them up to preparedness and then and then and then look at how they performed and then and then have the discussions around, you know, if they’re in the right place or not.
[00:26:15] So I think outside of the, you know, kind of AI. Algorithms and bias and conversational models and all those things. I think really the people and performance analytics pieces another area that we’re spending a lot of time on it and kind of excited to see that progress and I think it has the potential to really help offset some of you know, people just being put in positions and they weren’t they weren’t they weren’t the right fit for and that’s obviously not good for anybody.
[00:26:47] We should we should have another one of these conversations were that’s all we talked about right because as I listen to you as a person to their I thought oh, oh he’s talking about teaching the test. And that’s that. I think that’s an interesting 20th century view of how work is and how people are.
[00:27:07] But but I am profound persuaded that the work of the 21st century is going to be work that you can’t really understand the test unless it’s vocational stuff like it is becoming where it’s where it’s a mechanical technology that you do. But if if you’re in the if you’re in a job where you have to juggle machine output.
[00:27:33] And make up the difference between an 80% probability of being right and actually making a decision exactly what you don’t need to pass the test in the way that we that we’re currently doing in our Public Schools. So, that’d be great Irish Nation to get to well, I just stepped on that topic. I mean, I think.
[00:27:57] The then the notion of the test that you’re taking is actually even is even quietly potentially changing a lot. So so we’ve talked about very, you know, very direct examples of where I’m going to go into a job invite Beyond stimulate that job that actually job I’d become more prepared. There’s another layer of what’s going on here.
[00:28:21] You know, there’s there’s some movement obviously in the HR space especially where we’re people are using kind of neuro Neuroscience back games to assess deeper traits and natural abilities. And and all of that is, you know, when you when you take away to the interfaces and you put people back into a spatial reality of virtual or something where you’ve got your hands.
[00:28:44] You can move freely. You can solve problems that just the way you would if you didn’t have any barriers any motor cognitive barriers between you and and the screen that gets pretty interesting in terms of what types of let you know puzzles Puzzle games. Do we build to help assess things that are not, you know the direct practice?
[00:29:05] Because I think you’re right John. I think it’s getting a lot of this is going to become you know, it needs to be kind of a more a more core level assessment of mi mi really a good agile thinker. Where do I know if my desk becomes chaos? How long does it take me to become productive again? Right because those are the kinds of things that that obviously, you know, probably.
[00:29:32] Are going to more accurately predict somebody success when they’re going into some of these unknown spaces where work is changing so much. So so I mean it’s an area of extreme like passion for us that we’re trying to get to obviously we’re trying to walk people forward step by step into understanding all of these different benefits that come out of kind of this this new spatial Computing landscape.
[00:29:55] So but yeah, it’s a that’s a super interesting topic. Yeah. Well, well, let’s get to it. Next time Meanwhile, we’re at the close of our time together. Would you take a moment to reintroduce yourself? Tell people how they might get a hold of you. Yeah, so, my name’s Kyle Jackson. I’m the co-founder and CEO of of tailspin.
[00:30:18] You can find us at tailspin dot company spelled out on the web and on Twitter also had a tailspin company for our handle. Thanks Coco. I really appreciate you taking the time to do this. I look forward to spend a little bit more time with you later on. You yes think HR examiners executive conversations and we’ve been talking with Kyle Jackson who is leading the front end of the delivery of VR to HR applications in a company called tailspin.
[00:30:51] You can find them at tailspin dot Company, please look up Karl Jackson. You’re going to be hearing lots about them in the future. Thanks for tuning in today. Thanks again for doing this Kyle, and we will see you here, same time two weeks from now, this time next week will be coming live from the HR Tech Conference and Stacey Harris and I’ll be giving you an update about what happened. Talk to you soon. Bye bye now.


Read previous post:
Legal Issues in AI: Discrimination and Bias

“Bias is always an issue with AI because the machine learning systems only know what they are taught. And what...