Psychology Meets Artificial Intelligence with Grace Norris and Joe Meyer
Psychology and AI for HR.
"If you fail to invest in people and culture, then we end up with a very isolated, very individualistic dog eat dog kind of culture.” - Grace Norris
Join Grace Norris, a specialized Researcher & Data Annotator delving into the synergy between psychology and artificial intelligence, alongside Joe Meyer, an expert in Natural Language Processing and AI. Together, they will guide us through the exciting possibilities that emerge when the realms of psychology and artificial intelligence converge.
Subscribe to join our supportive community of people-first leaders!
Janine Ramirez: Hold on tight because it's a double dose of brilliance on today's special episode of the Employee Experience Experience, Erudite Edition. We've brought together two guests with two different perspectives, but united in one mission to revolutionize the world of EX., and that's employee experience, and you should know that by now. Okay, so now we have Grace Norris from Erudit Research Psychology team.
Her background is in individual differences and neurodevelopment psychology, with a first class honors in psychology with sociology. So she's experience in youth work and in education. So she will ground our conversation today into everyday realities of work and life. Hello, Grace. How you feeling?
Grace Norris: I'm good, thanks. Nice to be here today.
Janine Ramirez: Great. Okay. On the other side, we have Joe Meyer on board, a powerhouse in the fields of AI technology and organizational psychology. Joe is a senior natural language processing researcher at Erudit, and his expertise is in data science and natural language processing or NLP.
Hey Joe, how are you?
Joe Meyer: Hey, I'm doing well. Thanks for having me today.
Janine Ramirez: Good and excited. I'm Janine The Awkward Podcast Host that's going to try to bridge these two minds and bring together some cohesive insights within 30 minutes.
So, are you guys ready? It's like a game show, right? Let's do it.
Okay, so I've spoken to some, like, remarkable people, leaders already in the past in the podcast and in previous webinars. And I've observed that a recurring theme is the significance of investing in improving the employee experience. So investing in wellbeing, diversity, engagement and things like that. They see that data in research consistently demonstrate that these investments will have a positive impact on the bottom line on the business outcomes. However, there are also stories and instances where businesses choose not to make this investment.
Investing in People and Culture
Janine Ramirez: So like during challenging economic times when they're budget cuts and stuff like that. Most of the time, the people and culture initiatives like take the first blow. So I'm curious to hear your thoughts.
Like, first, do you believe that organizations should invest in people and culture? And what are your top of mind data and studies that support that?
Maybe we can start with Greece.
Grace Norris: Yeah, I absolutely believe that organizations should invest in people and culture. A healthy workforce and positive culture ultimately supports any organization in attaining its goals. If you consider to be a company to be a system, then system psychology denotes that the system and its component parts is best understood through relationships with each other and other systems. So not in isolation.
If you fail to invest in people and culture, then we end up with a very isolated, very individualistic dog eat dog, kind of culture. And things are more likely to fail in times of like burnout and turnover. And current research at the Chatting Arrangement 2020 Engagement Study kind of supports this in highlighting that there are lots of positive consequences for investing in your people, including kind of positive financial outcomes, reduced turnover rates, fire retention rates, and reduced costs as well.
Janine Ramirez: Great, like the data is out there. You know, for investing in wellbeing and your people. Joe do you have anything to add due to the datasets?
Joe Meyer: Yeah, I think I think Grace put it pretty well. I guess a lot of organizations, sometimes stereotypically see kind of human resource functions as a cost center on which to turn and it's been turned around. But like I said, I think when you look at the research and studies and the data, you see that investing in people from different perspectives can kind of impact the bottom line of organizations.
So I think that's important to consider that really it's more peripheral and interact with the support is definitely there. So I think it's really important to consider especially if you think about kind of like different types of organizations, for example, like knowledge workers and things like that. There's going to be a lot of pretty strong consequences if there really isn't investment in people.
For example, you can have really the most talented, brilliant team, but if they don't have the resources that they need to succeed in their work, then things really aren't going to do well for the team. You could see people leaving the organization, you could see burnout starting to happen, leading to things like mistakes or poor outcomes. So yeah, overall, I think it's really important to consider that investment.
Janine Ramirez: And I guess we all know how difficult it is to find the right people for the job and then to say there's a lot of investment in time and resources that go into that. So once you find them, I mean, logically, you should invest in their experience and you know, their everyday kind of work environment.
But why do you think that even with all this data we're still hearing of, you know, like HR professionals struggling to get support and budgets for their initiatives?
What do you think Joe?
Joe Meyer: Yeah, sure. I can start with that one and then, Grace, if you have anything to add, feel free to jump in. Yeah, I think you made a good point with kind of I think you were kind of alluding to like the employee lifecycle and investment is so important throughout the entire thing. So like even before people join the organization, there should be a concerted effort determining who should we even reach out to, whether that's cold emailing or whether that's like targeted campaigns on LinkedIn or something like that.
And then from there, there should be investment in some cases. If you think about the side of industrial psychology where you're developing high quality assessments to determine who is likely to succeed with in this job. Even before someone joins an organization who is likely to stay with us. And then you even have kind of investment in determining, okay, well, we may have determined that this person can be good with an organization, but what team that you can join?
So there's a lot of research that we can look at that and big things. But yeah, I think in terms of your question about why don't people invest in this, I think it kind of ties to what I mentioned before that there's seemingly an indirect connection between directly investing in people. And then you may see a return on investment maybe three months or six months or maybe even a year later, because if you even think about a high quality onboarding, could even take six months in some cases.
So comparing that to something and that's probably more obvious, is a little bit more difficult to see the impact on bottom line or something like that. So I think that's largely why some people won't consider investing directly in people as opposed to something that seems to be more direct in terms of immediate impact.
Janine Ramirez: Right. It's not directly connected. So even in our brains, it's not thinking, you know, an immediate result of the investment.
Do you have anything to add to that, Grace?
Grace Norris: I think Joe summed it up pretty well, but yeah, just I think often people and employees are seen as an expendable, disposable. They suppose, you know, there's someone else who could take the job. You know, there was always this millions of people out there looking for work, but it really is about who is the best fit, not necessarily who's just got the best credentials.
And it is looking back to kind of people and culture, about a kind of alignment of values as well. And so if you invest in your people and culture and you know what they are, you identify them strongly in your company, then you can identify them in your potential staff, too. And, you know, clearest asset you're going to keep and that will benefit the company in the long run.
Bridge the Gap
Janine Ramirez: Okay. So there's like this gap. I don't know. It's hard, I guess, to get buy-in from leadership, from C levels, to really invest in these people and culture initiatives.
What do you think we need to bridge that gap and to really be able to make a solid case for investing in the employee experience?
Joe Meyer: I think a good start, I guess, on that one. I also wanted to add, and I think one difficulty is that if you think about even the definition of culture, it's a little bit diffuse and people probably have different perspectives on it. So depending on the definition, you know, you may have different approaches to analyzing it. But I think one of the biggest difficulties is that, and this ties to everyday too, is that it's pretty expensive to even gauge what culture is depending on your definition.
So if we kind of run with maybe, and some people will argue on this, that grace I don't know if you've come across like I think his name is Edgar Schein. Some people say it's like very deeply ingrained. Other people say it's more kind of observable things. So I think that's more on the climate end of things. But you know, without getting into that too much, it's really difficult to analyze it.
If you're looking at survey based information and you want to get a gauge what culture is, it may take months to a year to really be able to get a pulse on that. And then from there you have to, like you mentioned, you know, gather by and from leaders to even kind of agree what the assessment says and things like that.
So at the end of the day, you may have a year of extended effort or months on several teams trying to analyze that, several teams trying to communicate the results back to the organization. So from there, you're looking at a lot of resources spent. So it's kind of difficult to create an investment after that. So I think one big thing is the difficulty of analyzing culture.
Janine Ramirez: That makes so much sense. And it's scary to think that it'll take that long and that many resources. And what the the pandemic taught us was that these things they change like that, you know what I mean. I think your culture, the way people are feeling, the way people are working and change so quickly.
So, you know, the the data that was analyzed is probably moot by the end of the process, right?
Grace Norris: Yeah, that's part of the issue. There's the gap between when the research results were finally published and when the events actually occurred that impacted your stuff. Like so, by the time burnout is identified, then the, you know, staff members already burnt out and may be on sick leave and it's hard to identify the cause and situation. It's reactive rather than preventative.
So it's hard for C-level levels and other people to see the value of these insights that we have. And another issue is that gap between the kind of academic level, scientific level information and kind of the information for the layperson, you know, is not always easily understandable or digestible. So, it needs to be easily accessible, easily understood and kind of quickly digestible to make actual changes, you know, to take action on this information that we have.
Janine Ramirez: But that's like the power of AI too, right? I mean, to get all the data condensed and digested and gives us like what we need. And I was talking to Edie Goldberg and I was asking, how do you get? She asked this idea for the inside gig. And I'm like, is that even possible? And she goes so many things that were not possible in the past for HR and for people are absolutely possible now in the age of AI.
So this AI help fill this gap and I'm guessing it's a yes because that's her mission, right?
Maybe Joe shows that's true what I said.
Joe Meyer: Sure, yeah. I can kind of come at it from the AI perspective, for then I agree so, if you have any perspective on the psychology. But yeah, you know, like you said, I think has the strong potential to kind of close this gap like we talked about before. It may take months to get the results at that point. You may have strong events within the organization that kind of invalidate your results and then you're acting on information from the past so, it’s reactive, like Grace was saying.
But if you kind of think about the connection between AI and language, there's definitely a strong potential for this. If you kind of think about culture and how language may connect to that, language could definitely be an extension or an expression of culture. So based on your experience within the organization, that's going to influence your communication patterns.
For example, if you think about the variable of burnout that we're very interested in, if someone is stressed out at their job, that's likely going to influence their communication patterns. So AI has the ability to be able to pick up on that. And with the help of Grace and her team we're able to kind of train models to pick up on those communication patterns.
And I think one of the strongest benefits of AI in terms of assessing culture is the kind of lightweight nature to it. So as opposed to doing a survey administration within itself, maybe the last two weeks to a month with a bunch of reminders, I able to kind of quickly and accurately gather insights, essentially in real time, of course, depending on a lot of different factors so like how long it takes to make predictions and stuff like that.
But regardless, it's much more lightweight and one of the large benefits too, is that there's a lot of language within organizations. So and I kind of go back to this quote kind of frequently, but some people estimate that around 80% of data within organizations is unstructured.
And in order to analyze that dependent on the organization size that can take an incredibly long period of time. But AI is able to kind of sift through that vast amount of information and make sense out of the unstructured information very quickly without human fatigue. So in some cases, we're definitely looking at the scale of human years to digest the information that maybe just even let alone read it as opposed to make sense out of it compared to maybe day a day or hours or days for AI to make the same judgments.
So I think there's definitely a lot of strong potential in terms of analyzing culture.
How Erudit AI Works
Janine Ramirez: For the listeners so they're not aware of what Erudit does, and maybe we're speaking in a bubble. So our AI actually sifts through the business communications and just spits out all the juicy insights, like your engagement levels and burnout risk and employee satisfaction and stuff like that.
So it comes from like really I guess organic conversations, business conversations, but I want to play the devil's advocate and be like Grace as a psychologist, like, isn't even accurate?
Because sometimes it feels like magic, right? Like, oh, like we're all talking and collaborating and coordinating with each other. And then all of a sudden, from that, we get all these metrics.
Grace Norris: And it can seem like magic when you don't know the many steps between getting an end point, but there's a lot of work that goes into it, behind the scenes of the AI as well. And of course, we have like a team of professionals, experts and we have very strict validation and training processes with use of a proprietary annotation tool.
And then we released the models to production only if they achieve at least 80% accuracy on, without noisy data. And of course we aim to always reach 95% accuracy. So, you know, that's a 95% confidence level that you know, that this is accurate. But as we start out, we kind of continuously iterate between 80% and 95%. But that's also why we have a team of psychologists to have that human perspective training the model and analyzing the language and applying it to analyze the culture and the temperature, all of these different metrics that we're assessing.
So there's a thorough process that goes into it. Yeah, it's not just magic and wizardry.
Janine Ramirez: And a quick question. I should know this, right, because I work here. One day it will get better in time and with more data, right?
Like that's my understanding of AI and machine learning.
Right, Joe? I mean, it will get better.
Joe Meyer: Yeah. One of the most basic definitions of machine learning, I think I've seen is like learning from experience, but in this case is data. And I think one very important piece of this puzzle improving over time is also data diversity. So being able to have access to data across organizations, across industries, across cultures, because you kind of see these idiosyncrasies in the usage of language.
In order to train a good model, you kind of need to have that diversity. But it seems like as models are progressing, for example, with GPT, there's a lot of strong benefits in terms of helping your models become more generalizable. So being able to do well in new situations and a big part of that is having access to data, large amounts of data and diverse data.
Working with AI and NLP
Janine Ramirez: Like it. I don't know. It makes me really excited. I think we're going to go through it like later on about, you know, our mission, what we feel about it. But just quickly, I'd love to hear the perspective Grace and to know if you've worked in AI, with an AI or in an AI company before. Because I have not.
So Joe is like my teacher, Joe and Rick. So like what do you think in general about AI and NLP how will it change the way we like, understand and interact with fellow humans, other people?
Grace Norris: Well, I'm with you on this. This is my first job in AI as well. I became familiar with it on OC2 Psychology. I guess it's the first time I've worked with it. So I think it's the place to be right now. AI offer so many potentials, so it's really exciting and is very experimental as well to be working in this kind of area.
And in terms of what EX offers, I think just the potential insight into human cognition and the way the brain works and the way that we learn is kind of what I'm most interested in. I think understanding how we learn is one of the things that let me down my path to psychology.
So to be kind of on the brink of learning so much more about that through AI and understanding, seeing what's seeing how the neural networks work in like supervised learning and things like decision making and problem solving. And if we can map those onto the human brain and potentially learn a lot about what's going on up there. So that's what I'm interested in.
Janine Ramirez: I love that perspective because like live in Spain, I have a lot of people around me that they're like afraid of AI and have like that apocalyptic view of AI. And they always go like, Oh, it's going to, you know, like take over humanity or, you know, what's going to happen to humanity?
And in my head is like honestly working here I'm learning so much more about people than ever before, so thankful for it. Of course there's the fear of humans using AI, but that's another discussion.
Grace Norris: I think AI is as powerful as the hands of the people that it's in. Essentially, it's people behind it, training it, using it, applying it. So I think that's something that's important to remember when there's that moral panic around the dangers of AI that, you know, people close the race, not the technology itself.
Janine: Exactly. Do you have anything to add to that Joe or are you like up to here with AI apocalypse discussion?
Joe Meyer: Yeah, I guess I have a few thoughts on that. I guess when people think about like the connotations of AI, they typically think like what's the movie Ex Machina? I'm not sure how you pronounce it, but the kind of sentient humanoid robot. But, of course, that's an area of research within the AI if you kind of think about like the Tesla bot or something like that.
But for the most part, and especially what we're dealing with, we're dealing with algorithms and computation and we're trying to just basically use modern techniques to make really good predictions. So when you kind of think about the downfall of society, it's a little bit harder to connect that with most use cases of AI now.
But yeah, I think I guess some of it comes from just a knee jerk reaction to that kind of like the stereotypes of AI.
That being said, I think we're still in the midst of an extremely pivotal moment in terms of the progression of artificial intelligence. If you think about things like GPT and Lambda and especially the connection of GPT to certain tools. So it's actually absolutely a very strong disruptive force in our understanding of work. So it'll be interesting to see kind of as a society how we kind of flex to that.
And I guess ultimately I view it as like a major new technology or tool if you kind of compare it sort of to like a calculator. Previous to that, you know, we had to do calculations by hand, but now we're able to use this little machine and it's able to do it very quickly and accurately. And AI it's kind of similar to that.
But there's kind of a larger disruptive force because now the calculator's expanded to almost every industry. If you think about something like agriculture, I was watching a viral video about a tractor, like an automated tractor. I'm not sure if anyone was driving it where it could go through a field and it could use computer vision to identify weeds or something of that nature and then point a laser at those weeds and just eliminate them automatically.
Janine Ramirez: I’m getting Black Mirror vibes.
Joe Meyer: Yeah, I don't know if I've seen a Black Mirror like that.
Janine Ramirez: I’m going to link to you that exact episode.
Joe Meyer: Love black marigolds.
AI for HR
Janine: And you seeing Black Mirror in a whole new light now after working in AI, but that's another discussion. I don't want to take more of your time because we have to do in this start up.
So I just want to end with like one last question for the two of you. Like, what attracted you and what is getting you excited over the field of AI for HR?
Grace Norris: For me it was the potential to apply like psychological knowledge, especially to real life change to actually make positive impacts in people's lives on a day to day basis, especially with things like burnout and kind of how prevalent that's become. You know, it was a little known concept not that long ago, but it's a real life issue and crises in many cases right now.
And so the ability to use knowledge and like combine, you know, things like AI and psychology. Just make people's lives better and improve working lives as well which you know, it's hard to get a job you love sometimes, you know, or work for a company that's going to invest to say if we can get this technology together. And use it to invest in people and make their lives better through as we work as well, then yeah, that's kind of what's out. Yeah, what brought me to it.
Janine Ramirez: I love that. Joe?
Joe Meyer: For me, I guess I studied AI and psychology and then kind of at the end of it got really interested in data science. So that kind of progression made a lot of sense. And I think one of the most exciting things to me is that compared to other fields, AI hasn't been adopted as much within HR so I think there's a lot of room for improvement as well as new innovations in the space.
So I think that's what I'm most excited for that mainly gravitated to this.
Janine Ramirez: I love it. Thanks for bringing out your savvy side, Joe and Grace. Thank you for taking the time to speak with me. This is really kind of an exercise also of internal communication and getting to know each other. I mean, we're working in like 100% remote setup, so it's hard to get to know everyone, right?
So thank you so much. I'm happy we got to do this.
I hope we can do this again another time.