HR Tech Weekly Logo

Hosts Stacey Harris and John Sumser discuss important news and topics in recruiting and HR technology. Listen live every Thursday or catch up on full episodes with transcriptions here.

HR Tech Weekly

Episode: 278
Air Date: August 6, 2020





Important: Our transcripts at HRExaminer are AI-powered (and fairly accurate) but there are still instances where the robots get confused (or extremely confused) and make errors. Please expect some inaccuracies as you read through the text of this conversation and let us know if you find something wrong and we’ll get it fixed right away. Thank you for your understanding.

SPEAKERS: Stacey Harris and John Sumser

John Sumser 0:14
Good morning Welcome to HR Tech Weekly One Step Closer with Stacey Harris and John Sumser. Hey Stace, how are you?

Stacey Harris 0:21
I’m good. I’m doing really good. Dropped off my youngest son at his the new college apartments yesterday. So, I’m officially an empty nester and enjoying the sun here in North Carolina this week. And how about you, John? Is this week going well for you as well in California? You guys I know we’re still hunkering down but at least getting decent weather for it?

John Sumser 0:42
No, it’s hot. This is the hardest part of the lockdown yet. It’s too hot to go outside very much of the time. So, you get out in the evenings and get out in the early mornings, but beautiful and it is California.

So, what’s going on with you?

Stacey Harris 1:00
It’s been a busy couple of weeks working on wrapping up The Annual HR System Survey. So I’ve been spending a lot of time doing the deep level data cleaning that we generally do at this point in time. So we are, you know, removing or aggregating duplicates and removing data that just was not put in appropriately and making sure that you know, people sort of knew what they were talking about. So my eyes are a bit crossed from spreadsheet analysis.

So, I did take the day off yesterday, like I said, to move my son and that was an experience because it was a move in that had no connection with people like everything we did even getting the keys like droplets, faceless. Everything in the apartment has been pre cleaned, pre sanitized, as you would expect, but there was all these extra special, you know, regulations and rules around what they could do and not do and where they could go because of COVID. So that was quite an experience to sort of move someone across the country and not really talk to another human being while you’re doing it, other than your family. I’ve been spending a lot of time focusing on getting my son settled for his probably mostly online, but the university here at least is going to require the kids to go to on campus a couple days a week. So we will see how that goes, and how that impacts COVID numbers here in North Carolina, and I have been feverishly warning my child to wear his masks and make sure that he case x rayed apart from everyone will hopefully he will listen like everyone else is hoping their kids will listen. So it’s been a busy busy week. But the new peer nature tech, interesting news, but not a ton of it going on right now. We’ve got a lot of stuff about announcements of people changing new role, which is exciting and new stuff. We’ve got some interesting stuff going on with what’s going on in the I would say the privacy space, both bills that are being put forward as well as technology that’s coming out now around the COVID work and a lot of really interesting I think conversations about when you reopen your organization, what kind of tech could make a difference, this new future of work, so not busy but interesting conversations. How about you? Anything going on in your current work that’s that’s worth having a conversation about?

John Sumser 3:07
You know, I talked to, I talked to an interesting vendor who will remain nameless because I’m in a better mood than usual, yesterday, that does video interviewing with AI. And we talked for a while, it was a surprising conversation to me. The idea was that they bought a $60,000 natural language processing module. And they can use this inexpensive tool to analyze language patterns. And what they try to do is match their predictive elements to the historical performance of the recruiters in the company. So this is for high volume stuff, and they can match what the recruiters have done in the past pretty closely. And so I said, well, what about the bias that introduces into the system? And he said, “That’s not our job.” Our job is to match what they already do. And bias is their problem. And I thought to myself, well, that’s kind of an extreme position. But maybe that’s what tech vendors have to do is say, it’s the client who is responsible for this stuff. And so we really don’t know anything about it. And so rather than everybody I’ve talked to about video interviewing has talked about bias and its reduction in one way or another. And this little company with its little cheap AI implementation is making a clear pitch that says, “hey, we can give you the data, but you’re in charge of fixing your own problems.” And I’m still puzzled by it. I’m still puzzled by it. I can’t imagine that you can get away with that, but I couldn’t find an arguement that said that was a flawed approach.

Stacey Harris 4:59
I will have to say I mean, when you think about it? It’s the same argument we get with gun control or speeding issues or I mean, any technology, right? There is a level of well, you know, the technology is is giving you something, it’s what you do with it. That is the is the real, if you want to call it crime, or the real action, oh, boy, does that put the companies on the line in a way that they are, I think are really unaware of in some cases? I’m not sure. I think the reason that I would push back a lot on that is unlike maybe other technology and actually have this conversation over the weekend, we were talking with a friend of mine, and he works at a large technology company here in the local area. And we were talking about artificial intelligence. Now, the biggest issue is you don’t know what’s inside those algorithms, you know, and once the system starts recalibrating itself, he said it’s really hard to see how those decision matrixes are being made. And that becomes the hard part. So yes, you’re giving me something But is the professional making decision based on that information? Or is the professional taking that data and doing something with it? I can’t give a clear answer as to where it came from. And is that really the bigger issue? We don’t know where you’re buying your food from kind of thing.

John Sumser 6:16
So at the heart of this, at the heart of this is a tough reality for buyers of intelligent tools. The tough reality for buyers of intelligent tools is that no vendor is going to accept the liability for the bias that you caused, or the bias that they cause. They simply aren’t. They simply have their boards of directors wouldn’t allow them to write contracts that had they’ve accepted that liability. So when you use intelligent tools, the risk is all yours. Right? This is the thing that I’ve been talking about for years now that you can’t turn around and go after the vendor if you don’t like the results is intelligent tool. Because the contracts will say, You’re responsible for your results that the AI is just an input into decision making. And if you make the decision and it’s a bad decision, it’s your problem not technology’s problem. And I think that there’s good principle behind that. If you’re driving down the highway and your speedometer says 60 miles an hour, and the cop clock hit 90, you don’t get a discount because your speedometer wasn’t working, right. So the tools that you use to measure your own performance, don’t have anything to do with the way the world perceives and reacts to your performance and you can’t expect to be protected from your own stupidity. Even if the AI has led you into it, you still can’t expect to be protected from your own stupidity.

Stacey Harris 7:51
I’m seeing John’s t-shirt now you can’t be protected from your own stupidity.

John Sumser 7:56
Oh, that’s what we need is an HR Tech Weekly Merch shop.

Stacey Harris 8:03
But that is a, quote, very quotable quote. But yes, this is exactly it, that the machine does not absolve you from making a decision that is ethical, correct or right. And I think this is the reason why any kind of technology requires any kind of data set that you’re using requires multiple bench analysis. You know, we were talking about this the other day, like when you do compensation analysis, you would very rarely ever use just a single data point. At the very least, if you’re doing really major compensation analysis, you would use at least three data points to just see, because you’re very aware that one might be out of whack, but there might be things that are biasing the compensation number and that one particular data set. So you want to see other data sets. And I don’t know that we don’t get that conversation in these individual tools that we’re going to show you different data sets, the conversation is our data set or our algorithm is the bat and you should use it and it’s really up to the businesses to day to go out and find other data sets that might either validate or negate what those data sets are saying. Correct?

John Sumser 9:06
Yep. And so in the case of this vendor that I was talking to, it’s not possible that they understand the problem that they’re trying to solve. They haven’t invested enough money to be smart about it. And right, because being smart about these topics requires making a lot of mistakes. And making a lot of mistakes is another word for r&d. That’s what research or development is all about is the concentrated making mistakes, and on a tiny budget implementing somebody else’s algorithm without some sort of check, but you can’t do that. And so the conversation with a vendor illuminated a bunch of things like how important it is that when you acquire some sort of intelligent tools, you dig deeply into how they’re constructed. Because I saw this great thing yesterday, there’s a black hat hacker. conference going on this week. And at the blackhat hackers conference, this 17 year old kid talks about his hundred dollar project and with this hundred dollar project, he was able to make a working version of Tom Hanks that looked and moved like Tom Hanks and said things like Tom Hanks, and it was assembled from audio clips and still pictures, wow, hundred bucks to fake Tom Hanks. And he couldn’t figure it out on a movie screen, but he can certainly fake it in a web delivery. And that’s how the modular components that are available in open source or by vendors are and if you don’t have a clear understanding of what those components are, what is your new system built out? Have you set yourself up for being surprised by the fact that it produces Evidence of your own incompetence, or introduces more incompetence into your system. What do you think about changing the word bias to incompetence? Would that make that conversation easier?

Stacey Harris 11:12
It would maybe get a better reaction as far as we need to take action. Because I think we say the word bias now and everyone kind of thinks, Oh, yes, you know, I can’t do anything about this it’s in everything we do. Right? We know it’s an issue, but it’s always we know, it’s an issue. But I think if you say incompetence, yeah, you need to take action against incompetence, it’s not something that you can not change. So yeah, that might actually change the dialogue a little bit because it really is the inability to, and I think you really made that a great statement and the fact that this is all about understanding what’s inside of it, but also realizing that you have to keep on top of it, keep changing and keep looking at it. Right. Well, we’ve got a lot of other things that are I think you’re going to lead right into this conversation. There’s a great conversation Now about natural language processing that I think exactly that same issue as well as now we’re running into beyond just making decisions about recruiting and hiring. We’re adding biometric data to artificial intelligence analysis. And now we’re making decisions about people’s health, which puts a whole different level of concern, I guess, I would say to this conversation. Have you started to even think about it beyond beyond this area?

John Sumser 12:28
No. No. I haven’t even a little bit. What are you thinking?

Stacey Harris 12:33
Well, I mean, some of the stuff that’s coming up this week, you know, we’ve got a really interesting stuff coming out from a company called Brainworks, which is launching a medio smart health monitoring app, which is a free web app that uses AI enhance non contact measurement of your vital signs. Basically, it’s using facial recognition tools, facials for monitoring tools, digital health care, and to basically give you a readout of whether or not you might have symptoms of PCOS. And some other things that are going on, right, just based off of how many times you take a breath and how often Your face is moving. I was reading this and was fascinated. So what you’re talking about when you talk about the idea of bias being built in, and now you lend that to this next level of conversation, which is whether or not someone in the organization could possibly be symptomatic for COVID, or have other health problems. And can you make decisions on that? Should you make decisions on that? What does that really look like? This gets really scary really fast?

John Sumser 13:30
Well, you know, the job of HR is changing as we speak. And where the primary focus was a combination of talent management and acquisition and benefits in compensation administration and policing of conflict. There’s this new thing. It’s going to be inescapable that HR departments are responsible for health, both of the organization and individuals. ization. And there are all sorts of tools about it. I was talking to a doctor yesterday, I have a home oximeter that changed my blood oxygen levels. I was telling her about my scores. And she said, Oh, I wouldn’t worry about that, because the home stuff is not calibrated very well. Yeah. So if you’re generally in the range, it’s generally right. But it isn’t accurate. And so the question about what would have been seen as invasive monitoring devices in the office six months ago, is probably how do you know how accurate they are? And again, this is that line where bias and competence are related to each other. And the bias in a given measurement device is also as easily understood as the calibration of the device. How accurate is it? And so we don’t know. Nobody is testing this stuff for diversity related impacts. Before they ship it to the marketplace. So I don’t know where I got this, but I’m pretty sure that generally speaking, men and women have different baseline body temperatures, for instance. And so if the machine that takes your temperature from across the room doesn’t account for a gender variation, and I don’t know what else the things are that cause temperature variation, but if it doesn’t account for it, then it’s going to introduce bias into the system. And we’ve got an avalanche of stuff that measures monitors and reports on people coming in from bathroom cleaning robots, two sensors that measure whether or not everybody in the office is six feet away from everybody else in the office.

Stacey Harris 15:46
You had shared the CB insights report, which often does really good research on the tech enabled office in a post COVID world environment and the things that are in this picture, autonomous cleaning solutions, facial intelligence, people counting sensors, I mean all the things you talk about. Cleaning of the bathrooms and monitoring the bathrooms in a way that’s I guess appropriate. Wow. Still think we’re in a whole different world, but they’re here. And they’re being used in a lot of organizations already.

John Sumser 16:12
Yeah, it’s crazy. And you know, just to hone in on the bathroom question. This is the primary disease vector in the office. And so figuring out how to rearrange bathrooms space so that it’s safe, that you could do social distancing is, I think, a massive problem that people have totally underrated because right now bathroom design this, how close can you get people together without making them feel like they’re together and the bathroom design of tomorrow is how do you make sure everybody is six feet apart and that all of the droplets in the air are managed, very different problems. And so my guess is we’re going to see a lot of bathroom renovations go on in office buildings

Stacey Harris 16:58
And you know we have conversation about who owns the information? And there’s no doubt that this is oftentimes if you have like an environmental engineer, you have someone who has an office, an office manager or a building manager, facilities, managers, these are the kind of things that they may be working on. But it’s my take on this as an HR own, the use the processes, they might own some of the purchasing. And they definitely own the continuous data that’s been gathered and how it’s being used, as well as the conversation around bias, and particularly today on the conversation around bias and all of these types of technologies. Do you disagree with that? Or do you think that that is the case, that HR’s going to own a big part of this?

John Sumser 17:38
Well, you know, HR owns health and medicine in different ways in different industries. So if there’s a doctor in the plant and in a lot of heavy industry in the north, where it’s possible to have your arm chopped off if you do the wrong thing. There are doctors on the staff who have offices of this Have the factory. And those people work for HR, the health and fitness of people so that you could tell whether or not they’re allowed to go on the shift. That’s HR. It just hasn’t been normal. Because, you know, what are the safety issues at Google there, whether or not you can tolerate the loud music, the worker next to you is playing. There’s not a ton of health and safety issues in high tech companies. But in lots of other places. It’s just normal for HR to own the health and safety thing. And so I think this falls squarely in HR’s lap. All the way.

Stacey Harris 18:37
It’ll be really interesting to see how we connect the dots between the more practical keeping our work environment safe to the more I think theoretical but still really important conversation about how we address biases because oftentimes, that is very much something that we talked about, but don’t do anything with and definitely how we think about ensuring that the data is being used ethically, because those Things are tightly connected. But I don’t think that a lot of people are realizing how much they’re going to be connected. You know, I want to make one note about some of the other things that are going on this week.

These kind of conversations are really critical. And our vendors are going to play a big role in how your organizations address these. I mean, as you mentioned, it was a vendor who kind of have this feeling like that’s not my job. I don’t think the big vendors have that perspective. And in the big vendors are pretty clear that they have at least some responsibility in this area. We have some pretty big announcements. This week, we had Gretchen Alarcon, who many of you people know has been really the leading voice for Oracle’s HCM application and their HCM application product. She’s been the group Vice President, and before that at Peoplesoft for over 15 plus years, she has just recently joined ServiceNow as the VP and general manager of the human resources service delivery business and really interesting to see that not only the shift in the market, but also the fact that ServiceNow which has been more of a helpdesk role and been more of a solution around your enterprise users interactions and data capturing is continuing to step into the HR technology phase as a central role, more and more over time. And you just saw this even as they changed over the years. And Bill McDermott just recently, at the same time, we also saw in a new role Jason Seiden that became Chief Marketing Officer of Moovila, a smaller organization, but I think along the same length organization that’s trying to change how we think about HR in hrs role in the organization. John, do you think that in both cases, these leaders and they’re really important leaders in our industry, in many cases, you know, the kind of thing they’re gonna have to deal with is stuff that we’ve just been talking about for the last 20 minutes?

John Sumser 20:44
That’s right. That’s part of it. The role in both ServiceNow and Moovila is to increase the quality of the interaction between people who are doing work and get rid of the hassle. 10 years ago, you would never have thought that that was part of HR. And in both cases, it’s not entirely the HR game. Moovila, Jason Seiden is going to be great in that job, but Moovila allows better communication between the people who are involved on the task and the project. So it’s a performance management alternative. But instead of getting scores so you could tell who gets a raise the performance that they’re concerned about is the performance of getting the project done correctly. While ServiceNow serves HR functionality. It’s all about making the interaction between the individual employee and the organization smoother so that they can go back to where to get work done. The emphasis now is quick answering of questions, right.

Stacey Harris 21:49
Quick answering of questions and capturing of the data of all the questions that are being asked. I think the two things that serves down has really put a stake in the ground is that yes, we will answer questions you can get back to work quickly. We will also So be the tool that captures things that we can serve to resolve the problems before they become question. I think that the other play they’re making right now that we are a tool that will allow you to get ahead of what could be problems in your organization, which is a very different take. But a very HR conversation.

John Sumser 22:16
It is a very HR relation. So I think what we’re seeing and COVID just accelerates this, what we’re seeing is a transformation of the HR department. And it is extensive transformation of the HR department that simultaneously pushes it into granular operational detail while being the driver behind looking at the entire organization from a big picture data point of view. And I don’t think I don’t think HR has experienced anything like this before. It’s a big transformation. And we’re just starting to see the edges of it.

Stacey Harris 22:57
I would have to agree. I think this is probably the biggest shift for HR since the days when they stepped out of being a personnel department into being a strategic partner, I think this is the next generation of what HR will become. And it is much more encompassing. And I think, you know, it’s interesting, the quote that they gave on Gretchen starting, Gretchen is going to paint a bigger picture and help take the employee experience to the next level, HR’s role is much bigger than I think we’ve been willing to have a conversation about. And because of that, it’s going to be really interesting to see how the technology ensures that they do that. And if we can ensure that technology’s not stuck in mired in bias and diversity issues and algorithms that nobody can read. So it doesn’t alleviate the problems.

John Sumser 23:41
I think it’s gonna be interesting to see how this impacts your work with IRIM. I think that’s gonna be a fascinating edge. Because when you talk about what HR needs to do going forward, we can’t even yet define what the skills are that you need to be part of that change, but they’re going to be more technical than not, and that means that the role for the HR technology professional association is going to grow over the next five years.

Stacey Harris 24:10
We just did an introduction to HR technology, a six part session that I did for IRIM. And one of the big things that came out of it was what were the skills that people were looking for going forward? And you would have spelled this out because one of the things that people were asking about what more skills would you like sessions on? How do I review and understand algorithm as I’m making decisions on the HR technology so they get it? The practitioners understand that these things are changing rapidly. How we get them those skills. That’s a whole other level of conversation.

John Sumser 24:40
Yep. Yes, it is. Well, what a great conversation today Stacey and congratulations, Gretchen and Jason.

Stacey Harris 24:49
Definitely for both of them.

John Sumser 24:49
Thanks for doing this again Stacey. Another great conversation and thanks, everybody for listening in. This has been HR Tech Weekly, One Step Closer with Stacey Harris and John Sumser. We’ll see you back here next week. Bye bye now.

Stacey Harris 25:02
Thanks everyone. Bye!

Read previous post:
Talent Mobility – The Struggle Is Real and AI Can Help

Outdated personnel information is a key roadblock to realizing the next leap in productivity. Fara Rives from HiringSolved looks at...