Ep 16: Navigating the Skills Landscape for the Evolution of Work with Gordon Ritchie

[00:00:00] Hey you with the podcast in your ear!

[00:00:02] Just a moment!

[00:00:03] Have you already activated your mobile happy hour?

[00:00:05] So you have an unlimited data volume every day for at least one hour.

[00:00:10] It's quite new and for everyone who now gets mobile phone plus internet for at home on the TV.

[00:00:16] Now Magenta 1 customers will activate the mobile happy hour in the My Magenta app

[00:00:21] and it's already off, on the 5G network of Telekom.

[00:00:24] So don't wait, now get in and activate the My Magenta app of Telekom.

[00:00:40] Hey it's Bob Holver.

[00:00:41] In this thought provoking episode I sat down with my friend Gordon Ritchie,

[00:00:45] Principal Consultant at Skill Collective and my go-to expert for anything related to skills architecture, strategy and intelligence.

[00:00:52] Gordon and I explore the complex landscape of skills in the modern workplace.

[00:00:56] We delve into crucial topics such as skills assessments, ownership, taxonomy and the role of AI in skills inference.

[00:01:02] Conversation also covers the challenges of skills based hiring,

[00:01:06] the impact of automation on talent acquisition and the importance of internal mobility and reskilling.

[00:01:11] Join us as we navigate the intricate world of workplace skills and discuss strategies for continuous learning

[00:01:17] and adaptation in an AI driven era.

[00:01:20] Hope you enjoy the discussion and thanks for tuning in.

[00:01:24] Hello everyone, welcome to another episode of Elevate your AIQ.

[00:01:29] I'm your host Bob Holver and with me today is my friend Gordon Ritchie.

[00:01:33] How are you doing Gordon?

[00:01:34] I'm great thanks Bob, great to be here.

[00:01:36] Thank you so much for spending some time with us.

[00:01:39] You are the perfect person to talk about skills with, skills is your middle name.

[00:01:45] Yeah, Skillosophers my LinkedIn handle.

[00:01:48] There you go.

[00:01:50] If you would just give us a quick little background about how you wound up where you are in terms of,

[00:01:57] well you can talk about location changes and your moves geographically as well as your education background

[00:02:05] and where you've been working.

[00:02:07] Yeah.

[00:02:08] Solving these skills problems for everybody.

[00:02:10] The jungle gym journey as people talk about.

[00:02:15] So it's like my first engagement with kind of skills and assessments was in my early 20s

[00:02:20] when I was struggling to figure out what I was going to do and my dad paid for me to go to

[00:02:25] what was an office in Baker Street in London which was where I was living

[00:02:32] and sit down and do a battery of tests with a psychological assessment and meet with us

[00:02:39] psychiatrists afterwards.

[00:02:40] It's like okay this is what your aptitude and capabilities should be Gordon and it was

[00:02:45] nothing to do with where I was.

[00:02:48] So a very early career change driven by assessment around skills and competencies and

[00:02:55] the rest as one might say is history of kind of following that through

[00:03:00] enterprise learning for a big chunk of my career in the UK before I moved to the US

[00:03:06] out here in Washington when I started a technical writing company developing skills for

[00:03:12] people at large and through both online and classroom learning and then

[00:03:19] got into the skills, the more specific skills side of it with a very early pioneer in

[00:03:25] skills, inventory, assessment and analysis.

[00:03:29] Whilst I was in the US it was actually based out of the UK near Oxford called Infobasis and

[00:03:36] we were doing skills taxonomies in spreadsheets and building them that really seemed like a great

[00:03:44] idea and concept and that's where my specific skills journey started and

[00:03:51] we moved from acquisition to salary.com to Conexa to IBM where I wound up ultimately owning the

[00:04:01] IBM talent frameworks as a skills taxonomy and dataset and very early some of the

[00:04:10] Watson AI applications of that kind of data into career coaching and counselling and

[00:04:18] applications there and how could we use that as a algorithm for that and from there

[00:04:28] consulting and some startups more recently around the whole AI labor market data model with Hitch

[00:04:36] that was recently acquired by ServiceNow as a talent marketplace and has been re-platformed into

[00:04:43] the ServiceNow employee experience engine and more recently with Skyhive which is now at Cornerstone

[00:04:53] and to Skill Collective where as a bunch of skills journeymen I suppose and geeks to boot

[00:05:05] we're a bit frustrated with the hype and the false promises and the hit the easy button

[00:05:13] and skills will make it all good so we're kind of pulling the curtain a little bit back and

[00:05:18] being shining a light on some of the challenges but really to help simplify the confusion

[00:05:25] chaos that's in the skills market and you know me I'm the absolute kind of advocate of skills

[00:05:33] still the potential is as real as it was 20 years ago when I kind of got into it and we're really

[00:05:40] close to being able to realize that but there's still there's still a ways to go

[00:05:46] but I'm in it for the ride I've tried to get away from it a couple of times it keeps

[00:05:50] booking me back in yeah no that doesn't surprise me I don't think some of these debates

[00:05:55] are going away and we definitely need your expertise here so Skill Collective I mean is it more

[00:06:02] like a consultancy and advisory or do you actually build some of the skills intelligence

[00:06:10] software or both? We're very much a consulting organization and basically providing

[00:06:21] three things to both sides of the coin right we've poacher turned gamekeeper almost we've been on

[00:06:27] the enterprise side we've been on the vendor side we know the mismatch of conversations and the

[00:06:33] fears of buying the wrong thing that's on the enterprise side so we're really providing a kind

[00:06:39] easily accessible expertise for enterprise okay what's the real truth behind this question

[00:06:46] Gordon we're getting this insight or for vendors it's like okay we want to put the

[00:06:53] skills sticker on our box but we don't want to just do the low-hanging fruit MVP

[00:06:59] 101 list of nouns what's the what is the white space or what is the leapfrog motion

[00:07:07] that education and enablement and if centers of excellence or expertise need or commenting

[00:07:15] become part of that team and knowledge share that upskill that within an organization

[00:07:22] so that they can take it and run with it ultimately everybody has to have their own

[00:07:26] sense of ownership around it otherwise if well when the one advocate leaves

[00:07:33] the project will stall we've seen that so many times. So when it comes to ownership

[00:07:40] is it learning and development that kind of owns this is it the CHRO is it maybe it depends right

[00:07:50] but it just seems like with the disconnect I don't see consolidated ownership of

[00:07:58] talent and talent pools within an organization and so it's hard to be consistent in how you

[00:08:07] view and assess skills if you don't have that shared view of these different talent pools and when I say

[00:08:17] different talent pools I'm talking about your current employee population your contract and

[00:08:24] you know contingent labor and then just you know people you're trying to source to target for

[00:08:31] you know new candidates right if there's a firewall between you know TA and TM for example

[00:08:38] it's going to be hard to do a lot of things right including strategic workforce planning and

[00:08:42] lots of things. Yeah the fiefdoms of those kind of HR vertical practices is one of the challenges

[00:08:50] for skills right it's like who does own it and it it shows up in the vendor space too because

[00:08:58] all of the solution providers to each of those verticals wants to be the system of record of skills

[00:09:03] and it's like why you just just put it all in the hiring system or put it all in the HRAS or put

[00:09:08] it all in the LMS and that'll be the system of record and each of those needs different data

[00:09:15] elements for it but who really what is the real invested interest in having that I've been

[00:09:24] fond of saying a lot recently it's like the skills that got us here won't necessarily get

[00:09:30] us to where we want to be I think that skills are not the end game so if all we're doing is trying

[00:09:37] to centralize skills a bunch of data elements that's not going to get us to the holy grail

[00:09:47] which is the business of the business right that's where it needs to be owned and have a

[00:09:55] sense of engagement with if it becomes a its own HR process that falls into the done to you

[00:10:03] not for you as employees and there's no connection to the tasks and the work that you have to do

[00:10:11] we'll be having this conversation in another five years like who owns it where does it sit

[00:10:16] I feel like we've seen this a little bit before with with DEI right we had an office of DEI we had a

[00:10:24] central kind of chief diversity officer and so we had the kind of vertical stack of paying

[00:10:31] attention to it and that checked all the boxes but it didn't blow out necessarily and the last

[00:10:39] year or two we've seen a lot of those diversity officers but the practice of that

[00:10:47] loose funding and disappear and not be invested in us as they could be because it was verticalized

[00:10:56] skills is broader than that and what it's actually about the 97 percent of the organization

[00:11:02] is focused on doing it's not for the 3 percent who sit in HR want to measure themselves against

[00:11:08] well DEI has its own challenges right it had a branding problem it had a it has legislation

[00:11:16] and risk associated with it and in hindsight there were better ways to approach that because

[00:11:22] ideally you know it's just part of business as usual operating procedure anyway yes it should be

[00:11:30] and so I mean even AI some of the premises behind using AI assuming you're using it

[00:11:37] responsibly and fairly which I certainly advocate for but in that case it seems like if you design

[00:11:44] and build and use AI responsibly when it comes to assessing talent you would naturally have the

[00:11:54] diversity that you were looking for at least over time you would right because you'd obfuscate

[00:11:59] you know certain fields when people are doing evaluations once you've mitigated the bias within

[00:12:05] AI the results and the insights would show you how to further mitigate the human bias that has

[00:12:13] pre-existed since the dawn of time on this on the skill side I suppose there is some

[00:12:20] bias towards you know looking at historical data well these are the skills that people had

[00:12:26] who were successful before therefore we need to find more people with those skills so that they

[00:12:32] will be successful too probably some hypotheses that need to be retested there or just acknowledge the

[00:12:41] fact that the nature of work is changing and like you said what got us here or what got us

[00:12:47] success in the past won't necessarily be what gets us success you know moving forward I think

[00:12:54] that's always a danger but looking at someone's maybe prior accomplishments had a different you

[00:12:58] know not just a different organization but maybe a different type of organization different type of

[00:13:02] culture and as far as people that are trying to start with like a skills inventory and you know

[00:13:11] building the skills taxonomy is that a waste of time skills are constantly changing skills needs

[00:13:20] half-life of skills all of that is changing constantly it just seems like that would be a

[00:13:27] never-ending exercise to take your skills well at least to try to build like a rigid sort of taxonomy

[00:13:34] for skills because you're always gonna need constant you know curation yeah I mean there's

[00:13:40] several things in that there's like just because it's hard doesn't mean it shouldn't be done

[00:13:45] right the one ring to rule them all kind of skills taxonomy or global skills standard or

[00:13:54] rosetta stone or whatever it gets labeled as right I think that's a little bit of a fool's

[00:14:00] error I think it's a large fools errand actually and I've seen seen and been part of a lot of

[00:14:07] effort to try and create that I think when it comes to like skills as the currency of work

[00:14:14] which is one of the other things that gets trotted out the way that I think about it is

[00:14:20] we're all our own fiat currency right based on our experiences and who we are we're all

[00:14:28] different but what we need to figure out is it's like the foreign exchange marketplace so that

[00:14:34] we can trade equitably what I want to do with what I'm able to do with somebody else who needs

[00:14:41] it doing for them I may have skills that a bank might pay a lot of money for and a non-profit

[00:14:50] can't afford that but I'd rather go deploy them at the non-profit but that's my fiat currency work

[00:14:58] that's my choice to a certain degree and trying to make us all the dollar or all the euro or all

[00:15:05] the pound or whatever doesn't work I fill up my car in Spokane and I drive to Seattle and what I pay

[00:15:12] for a tank of gas here in Spokane is different in Seattle but it's the same currency I haven't

[00:15:19] even I haven't even left the state and it has different value and worth and so we need to

[00:15:25] be able to think about that in the context of the skills marketplace the half life of skills

[00:15:32] drives me insane as a as a kind of non-statistic you can't find the data to back it up it just gets

[00:15:42] used as a kind of truth and part of that I think is the skill is part of what of the current skills

[00:15:52] that won't get us where we need to be agile software but think about it has forced us to release

[00:15:59] lots of versions of applications ever more quickly that inner skills taxonomy

[00:16:05] that's only listing product names masquerading as skills has python 56789

[00:16:15] and those inner data science mode unless you're being very sophisticated would show up as unique

[00:16:22] skills and look like they're just before we move on I need to let you know about my friend Mark

[00:16:28] Pfeffer and his show people tech if you're looking for the latest on product development marketing

[00:16:35] funding big deals happening in talent acquisition hr hcm that's the show you need to listen to go to

[00:16:44] the work to find network search up people tech mark Pfeffer you can find them anywhere

[00:16:51] five is one skill and python six is another skill but the delta is tiny and their increments

[00:16:58] of the same thing and so python five has not been invalidated by the invention of version six

[00:17:06] that's part of where the half life myth gets created because it's too simply calculated

[00:17:13] of skills counts when it shows up in job posts in the labor market well python five is now old

[00:17:20] and gone because python six is the thing that's getting listed but it's dependent on the

[00:17:26] foundational skillset knowledge of the previous of its predecessor yeah when I think of skills inference

[00:17:32] I think of of two things one is can I tell by the way that someone described something even if they

[00:17:40] didn't explicitly say they have that skill I can tell they have that skill by the way they

[00:17:45] describe whatever it was that they described but I thought the secondary sort of value of skills

[00:17:54] inference was to detect exactly what you just described like to know that there is a two month

[00:18:02] difference between someone who's on you know who knows level five or version five before they

[00:18:10] could upskill to be version six therefore if you have any rule whatsoever that says if this person

[00:18:17] doesn't have level six or version six then they should be eliminated from contention for a particular

[00:18:26] role or assignment then that's just terrible basically so I think about it is that part of what

[00:18:33] skills inference attempts to do or is that just something else that any intelligent you know

[00:18:39] skills solution would be able to account for yeah it's a great question I mean time is a terrible

[00:18:46] proxy for expertise because it depends on the actual application and so a lot of the inference

[00:18:55] engines will look at skills in a resume that have been in a job and the time that I was in that

[00:19:02] job for five years so I must be an expert as opposed to somebody who was in that job for two years

[00:19:07] may be less expert but one person could be repeating what they learned in the first six

[00:19:14] months over time and not really actually growing and this is another part of where the simplicity

[00:19:22] of the skill nouns struggles going forward the skills of the future will be about evidence

[00:19:30] a portfolio of evidence of the task tool of work that we've done it's the natural language of how

[00:19:39] we talk about work and we need to get to that point very quickly in my opinion because we have the

[00:19:46] capability with AI and the large language models to use the descriptions of tasks of

[00:19:53] responsibilities of what I've done to be able to infer the skills behind that it's not skills

[00:20:02] first skills underneath it's skills underwriting that and when we're able to engage in these tools

[00:20:11] that don't say well list your 50 skills Gordon show me the evidence or tell me the

[00:20:16] the tasks that you did today I know we've joked about it in the past nobody nobody goes

[00:20:22] to a party or goes home and responds to the question of what did you do today

[00:20:27] why did some critical thinking effective communication and problem solving level three

[00:20:32] we just don't we don't respond that way and I hope we never will text messages have gone from 128

[00:20:39] characters to unlimited and guess what we fill them up no end unless you've me and I'm still kind of

[00:20:46] constrained into 128 characters in my head which tries my wife and daughter nuts but it's like

[00:20:52] but we can do the natural language thing and that's about tasks and outcomes and evidence of

[00:20:59] the skills are behind that and that's where I think we can get to and should be yeah my first

[00:21:06] instinct when I started talking about some of these these technologies in this domain is to think of

[00:21:14] like hard and technical skills but when I think about AI and I see that the speed and

[00:21:21] trajectory of the capabilities that AI solutions are taking on I think of two things one is what you

[00:21:30] talked about before like the half-life of skills if we're talking about skills that AI is now

[00:21:36] taking on then maybe the concept of the skills half-life is perhaps more acute or legitimate

[00:21:43] this is something that we need to be aware of but but the bigger thing that comes to mind

[00:21:49] is the growing importance that if those other things are true how are we thinking about what

[00:21:57] we used to call soft skills so the human skills the power skills there's a lot of new adjectives

[00:22:02] that people are using but one of the one of the adjectives that I really like that first

[00:22:08] heard from our mutual friend Antonia was durable skills these are the durable

[00:22:14] skills for the reasons that I just cited that AI is going to come on it's going to take some

[00:22:20] away some of the more not just the mundane tasks but just the things that are you know repeatable

[00:22:28] the things that we wouldn't lose human centricity I suppose if we were to let AI do

[00:22:35] some of those things so that we could move on and do higher value tasks of which I think

[00:22:40] there are plenty there's plenty of work for us to do as we automate things and as we use AI

[00:22:47] you know as our you know co-pilot to help us make better decisions or what have you but

[00:22:53] how do you think about I guess the universe of skills cutting across potentially more durable

[00:22:59] skills as well as you know the technical skills and then the trajectory of AI I mean

[00:23:04] it's sort of this confluence of you know factors that makes your job I'm sure more

[00:23:09] complex than it already is yeah and I'm glad you I mean you mentioned the durable skills

[00:23:16] Matthew Daniel that Guild education was was a very early proponent of that phrase and so if

[00:23:24] you hadn't brought it up that's what I was going to lean on because I think nobody masters critical

[00:23:30] thinking if you are a critical thinker you are always challenging yourself and there's

[00:23:36] it's a it's a lifelong continuum of these skills it's not well I've hit level five let's bring on the

[00:23:43] next critical next one I don't need to do any more development of my problem solving skills

[00:23:49] and I think as we lean into or accept AI whether it's whether it shows up at work or we

[00:23:57] fall into a BYO AI kind of bring our own AI to work approach which I'm going to be interested

[00:24:04] to see how that evolves the durable skills what is a good question to prompt what is the critical

[00:24:12] analysis that you want it to generate something from and how do you engage in a conversation

[00:24:18] with an obtuse tool that doesn't recognize things in the graphics that it produces because of how

[00:24:26] they're technically created it doesn't I was getting into an argument with chat tpt the other

[00:24:33] day about what I forget which it might have been Microsoft about an image and it had misspelled

[00:24:41] a word in the in the graphic and I wanted it to correct the spelling but it doesn't know there's

[00:24:46] a word in there because it's just pixels it's just it's how it's developed and it just got

[00:24:51] worse and worse and worse so it's how do you deal with that using those durable skills to get

[00:24:58] the right things out of it and I mean when we were at IBM I've been spent a lot of

[00:25:04] time talking about augmented intelligence not artificial intelligence right I think about

[00:25:09] that a lot because I think that's quite prescient about the importance of those durable skills

[00:25:15] and and continuing to develop them yeah I I have had the same argument with chat tpt

[00:25:24] and probably at least one other generative AI tool what I don't understand is it knew where to create

[00:25:34] the pixels or the letters right you put the letters in the right order they're on the same

[00:25:41] horizontal plane it reads like a word and yet it couldn't even if you explicitly say here's

[00:25:50] what I want to do I want to use I want you put an a and then next to that put a t and then next

[00:25:55] so it understands this and it renders it so close and yet it just misses like it'll duplicate a

[00:26:04] letter or you know it missed a word or it most reason I use it to try to see what images

[00:26:10] would create for for my podcast and I put elevate your AI queue in quotes and so it got that

[00:26:18] except it did AI IQ so two a's two eyes and then but then it also at the bottom it had like a footnote

[00:26:25] and it said like podcast like I at no point did I say or did I imply that the word podcast should

[00:26:34] appear in the image that's just part of my prompt right so it's it's definitely far from perfect

[00:26:41] but that multimodal thing where you know letters inside images that is just that was just beyond

[00:26:47] frustrating so I digress I guess when I think about some of the ways in which we tackle this

[00:26:56] skills thing I mean one thing I wanted to just go back to my original question around like the

[00:27:01] ownership and people using different sort of environments to assess skills like somebody wrote

[00:27:10] the job description and then the candidate invariably used that job description and took

[00:27:15] the specific words and the specific skills from that and then put it in their resume and

[00:27:20] then that comes back and it gets evaluated and so you have this back and forth the sort of cat

[00:27:27] and mouse game kind of silly whatever I'll have my AI resume generator talk to your AI job post

[00:27:34] exactly exactly which will then be scored by another AI when it comes back in and then

[00:27:40] hopefully the human recruiters are looking at that and not taking it at base value again

[00:27:45] applying that those critical thinking skills to say you know did it overlook something what have

[00:27:51] you but then you get into the interview process you know you're going to wind up interviewing with

[00:27:58] probably several people none of which none of them actually wrote the JD even if they have

[00:28:05] in front of them and then they're thinking about skills in the context of the real role as they

[00:28:11] know it as opposed to how it's written in the JD and it just seems like how are we actually

[00:28:18] matching on skills like we're doing what an organization might call skills based hiring

[00:28:24] you just need a name for it I mean while they're still a candidate yes you already have

[00:28:30] inconsistency and the actual skills that you're looking for okay I was just talking about this

[00:28:37] the other day to somebody I mean skills based hiring is a great example to talk about

[00:28:44] the vast majority of skills based hiring today is actually just skills based sourcing

[00:28:49] it's not skills based hiring it does not track all the way through the interview process to the

[00:28:57] point of an offer letter being made to you to your point even if there's assessments

[00:29:05] through the process there's an interview and a human decision we've just been working with

[00:29:12] a large retailer who's been doing a lot of alternative skillifying of their job descriptions

[00:29:19] if you if you will so it's not just degrees okay what are the skills we need and what are

[00:29:25] as alternatives equal alternatives to a degree and they've spent a lot of time doing that it's all great

[00:29:33] and the net result when they look at who they actually hire for the job are degree holders

[00:29:40] because the differentiator for the interviewer or the hiring manager is the brand perception

[00:29:46] of the degree that somebody hold they haven't changed the practices the behaviors of

[00:29:53] the decision-making people yet but they've now they know that they're going to work on that

[00:30:01] and I think that's that's good data to have or we've done this and it didn't change

[00:30:08] I mean what they have done is got a whole bunch more candidates for the same number of positions

[00:30:15] to fill which looks like the talent acquisition process is being much more efficient you put

[00:30:21] a much bigger number on top right and you divide it by the same smaller number you're being more

[00:30:26] efficient with the decisions you're making and so that looks like a great vanity metric to measure

[00:30:31] the success of skills based hiring on but it hasn't actually changed the outcome and behavior

[00:30:36] yeah I guess I'm trying to think about that differently or if there's any kind of

[00:30:40] metric I mean I don't know if there's a metric that can be been certainly not directly

[00:30:43] and solely to talent acquisition in that regard but it is interesting to look back at the

[00:30:49] patterns and see you know who people wound up choosing and it was it because of an actual better

[00:30:57] skills match validated by human beings not just letting a particular matching engine

[00:31:03] tell you so but yeah and then how much did the degree or should the degree really

[00:31:08] matter does the does the recency of the degree matter yeah I mean there's different

[00:31:13] weightings that you can put on that a couple of other things you mentioned in the kind of

[00:31:19] the question setup was what the hiring manager or the functional kind of line managers the team

[00:31:26] managers are worried about I don't think they think about the skills that people need

[00:31:32] right they think about the job that the tasks that need to get done and we used to have

[00:31:39] I was joking with with our other mutual friend Brian Hackett the other day right we used to all

[00:31:45] be quite familiar with the time and motion study officer showing up and watching the job get done

[00:31:51] the process the output and improving the process grouping a bunch of those things together and

[00:31:59] calling that a that a job but they understood the process and the task at hand and for me

[00:32:07] that's where the skills conversation has to get back to almost as kind of let's get back to basics

[00:32:14] what is it that needs doing because that's the business of the business

[00:32:20] and that's what a job architecture needs to mirror or mimic and when you plug skills into

[00:32:27] that it has to support the business of the business and I think we run the risk of like

[00:32:34] running to skills based jobs of losing sight of the work at hand I just need these skills

[00:32:43] know what you need to do are these five things and show up with the skills to be able to do them

[00:32:50] and it could be Google's word processor or it could be Microsoft Word or whatever

[00:32:55] it's like simplistic tools but whatever it is you need the tools to be able to

[00:33:00] deliver the outcomes because that's what we're all here for nobody's here just for the benefit of

[00:33:07] our own health yeah I mean we're definitely more than just a collection of skills just like jobs

[00:33:13] are more than a collection of tasks but it feels like in some ways people are trying to do that

[00:33:20] sort of atomic you know breakdown because they can build a business case that way a business

[00:33:27] case for automation or in some cases AI because you don't automate jobs you automate tasks

[00:33:34] and so if this job encompasses you know these eight tasks and we can automate you know at least three

[00:33:42] or four of these tasks and we can just start making some uncomfortable decisions about the talent

[00:33:49] it's not exactly true I mean I don't know this is I guess it depends on the perspective of

[00:33:58] you know who's building the business case who's reviewing the business case but

[00:34:02] there's more than dollars and cents involved you know you can look for cost savings and cost

[00:34:12] avoidance but you can't ignore the impact to you know culture engagement retention and all these things

[00:34:22] and there's very significant figures associated with those things in fact probably more significant

[00:34:28] since labor is your most expensive expense so it just seems like some people are making some

[00:34:35] relatively short-sighted decisions if they go that route but I think that's the that's part of the

[00:34:42] kind of ethical can you should you side of this right and I think what's probably going to get

[00:34:50] more frightening before it gets less frightening is the the uncomfortable reality or knowledge

[00:35:00] workers to have to quantify the voodoo that you do has tasks even surreptitiously as prompts right if

[00:35:10] you're prompting chat tpt to do something that's a task that you're asking it to do that if chat tpt

[00:35:17] wasn't there you would be doing now is that something that can be repeated and therefore

[00:35:25] it's kind of removing the well I'm a knowledge worker I'm a white collar worker not a blue collar

[00:35:31] worker so my work is my brain is articulating the AI prompts the tasks that it can do for us or help

[00:35:40] us really what happened in the industrial revolution finally coming home to roost the

[00:35:47] knowledge work desk and that's going to be really uncomfortable but it's not it's a place where many

[00:35:54] people have lived for 150 years based on education or demographics or whatever it might be what 200

[00:36:01] years pick your pick your steam driven invention of how far back you want to go but right I didn't get

[00:36:09] involved in that type of thing when I was at IBM but at NBC I certainly did because I was helping

[00:36:15] to build the automation strategy and it was all right let's let's start going through you know department

[00:36:22] by department identifying some of the use cases understanding what the lift is to actually you

[00:36:28] know get that done what's the ROI of that is it you know three three months six months etc and then

[00:36:35] and then what do you do with that so I wound up having to pitch this to the COO

[00:36:41] during like the strategic planning session and some of the senior executives that were sort of coaching

[00:36:48] me were like so you're showing that you can save you know let's just say you know 50 head count over

[00:36:53] two years and I'm like whoa I don't want anyone to lose their job there's a million things that we

[00:37:00] need to get done that these people are probably capable of doing if only they had the time

[00:37:06] and this EVP basically said you know I'm not telling you we're going to cut 50 people what I'm

[00:37:11] saying is you want to show the COO the savings in the head count equivalent savings take that win

[00:37:20] or anticipated win and then tell them tell them what to do with that savings obviously he's

[00:37:27] going to want to book some of the savings because he's a COO but propose part of your

[00:37:32] pitch should be how are you going to reinvest some portion of that savings and that could be

[00:37:39] however you've sort of built your case it could be maybe you want to hire some additional automation

[00:37:44] engineers maybe you need new software licenses and maybe you need to upskill or reskill some

[00:37:51] of those people because if you can save a couple thousand you know people hours a year

[00:37:57] then you can shrink potentially shrink some of these existing teams but like I said there's

[00:38:03] other work to do and then you get into like you mentioned before Hitch with you know the

[00:38:07] talent marketplace and internal mobility or whatever like there's plenty of other things

[00:38:11] that people can do and they have transferable skills that would enable you to actually

[00:38:16] move to another area and potentially have just as much success if not more

[00:38:22] if you knew enough about people's ambitions and trajectory of where those transferable skills and

[00:38:29] some of the specific domain skills could take them for another sort of win-win relationship with

[00:38:34] the organization so and those circumstances have occurred for a long time and it's been it's

[00:38:41] always been puzzling and the closer to the top of the kind of pyramid I've got that it's been

[00:38:47] interesting to understand some of those positioning and the kind of decision making around that

[00:38:53] we worked on a huge project at AT&T as they were transforming themselves from a wire company with

[00:39:00] a with a mobile problem to a data company with a kind of with a legacy wire problem how do they

[00:39:06] train how do they reskill 100 000 whole climbing wire man into the 21st century AT&T knowing full

[00:39:16] well that not all of them would want to we're going to make it but how do they move that because

[00:39:23] there there weren't 80 000 people sitting on the sidelines of the market but the labor market ready

[00:39:28] to do what AT&T needed them to do so how could we move that and just today I was reading about

[00:39:36] I guess the pre-eminent tax software company laying off 1800 people of which a thousand plus

[00:39:45] were sort of were not performing at the same time with a thousand open headcount and net new

[00:39:54] AI and developer positions and I'll speak for myself sometimes if I'm underperforming it's a

[00:40:01] disengagement issue it's not because I can't do the job but if I had the opportunity to go do

[00:40:06] something new like some of those under performers might have been great candidates for the net new

[00:40:12] positions at this software company that takes work to do and this the CFO says

[00:40:20] well we're going to do these two things unfortunately HR gets the rough end of the

[00:40:24] stick of having to execute on that saying hey actually can we spend a little bit more time

[00:40:29] and just trying to figure out how we can move some of these people because we're dealing with

[00:40:33] lives here and it puts HR and talent management teams in a tough place when financial engineering

[00:40:42] of the most expensive asset or I don't know if you can have an expensive asset it's like

[00:40:49] the cost center or it's an investment as an asset whichever way you want to think about it but

[00:40:56] yeah I I mean I've definitely been in that situation before and it wasn't you know it's

[00:41:06] unpleasant I mean the company's in such a financial conundrum that they have no choice but to let

[00:41:13] some people go then they're going to probably ignore what some other data might tell them anyway

[00:41:21] because they've got to cut somewhere and you know it might as well be here or whatever but technically

[00:41:26] on some level they they know better even if they're not really super forward thinking and

[00:41:32] they're not doing strategic workforce planning properly and they're not taking a sort of a systems

[00:41:38] thinking kind of approach to the organization so it sounds like you and I have both been the victims

[00:41:43] of that situation and we can sort of Monday morning quarterback it and be confident that

[00:41:49] we're thinking about it logically but financially it just seems like there's a lot of incentives

[00:41:56] that would need to change not to mention the investment in you know obtaining that additional

[00:42:03] intelligence because it's not like the data doesn't exist this isn't like some you know esoteric

[00:42:10] you know thing that it's all hypothetical or whatever no the the data is there you just

[00:42:15] haven't applied or invested in the the skills or the software that or both to help bring

[00:42:25] that to light because if you did you'd probably be able to do what we just told you to do

[00:42:30] which is shuffle the deck and then use similar yeah but different technology to find where

[00:42:37] you could have been saving in other places that haven't been getting as much you know scrutiny right

[00:42:42] so it's like you've got to look under the couch cushions sometimes as posts yeah it's another

[00:42:49] topic for another time perhaps but I think if we were able to book or if we're able to treat people

[00:42:55] as an asset which is the lip service that gets paid even though they're in the same breath they get

[00:43:01] listed as the most expensive item on the payroll if they're an asset why can't we book learning

[00:43:09] and development as capital expense as a capital investment in improving that asset instead

[00:43:16] it's operational expense and it's it's out the door there's no way to amortize the return on that

[00:43:24] investment in improving your assets which they get to do every day of the week on the buildings and

[00:43:32] infrastructure that really doesn't do anything it just sits there and degrades unless they spend

[00:43:39] money on it but they get to amortize it they get to treat it as a capital asset

[00:43:44] like I said the topic for another debate another time if we were able to make that one accounting

[00:43:50] change I think we would see a massive change in attitude to how we managed and what we were

[00:43:59] able to do with the people who worked for us there's a lot of things that happen as a result

[00:44:05] of the way accounting rules work right I mean just like when you know you're in a hiring

[00:44:11] freeze this happens all the time right you're in a hiring freeze you can't hire full-time people

[00:44:15] but you could spend as much money as you absolutely need to on contract labor yep right so

[00:44:21] contracting the people you've just laid off because you need the skill set but now you can pay

[00:44:25] them as contractors not as employees happens all the time right it's crazy you know I'll just

[00:44:31] harp for one more second on the internal mobility thing I mean it would be a huge oversight to think

[00:44:37] that people are leaving just because they found a job that pays you know a little bit more like if

[00:44:44] you weren't going to give them another opportunity they wouldn't stay and perhaps stay longer than

[00:44:50] they would have otherwise and be more engaged and loyal because you've never asked them what

[00:44:56] their ambitions are and if you did then as new opportunities would come up it's it's almost

[00:45:04] analogous in some ways on the talent acquisition side to candidate rediscovery like well they weren't

[00:45:11] good for this job you know just like they're not engaged or they're underperforming and their

[00:45:16] current the employee is underperforming in their current job but what if we knew exactly what

[00:45:22] they were looking for what they were capable of and then we positioned them such that we

[00:45:29] you know enable them to pull a 180 and all of a sudden their engagement spikes and they're

[00:45:36] happier and yeah I mean it could make a world of difference and so I mean I went through that

[00:45:43] again at NBC through that automation strategy I said well what jobs and what people might be most

[00:45:51] susceptible if they were to let people go as a result of automation well let's start with

[00:45:57] that population in terms of who we could target for potential rescaling you know and then say okay well

[00:46:03] have you ever thought about a more technical career and then give them some type of assessment

[00:46:07] to see their potential to make that make that switch because otherwise they see a dead end

[00:46:13] for their their role because it can be largely automated you know they see the writing on

[00:46:19] the wall and if it's not this round maybe it's going to be the next round in three months

[00:46:23] or whatever and so and they just feel stuck and so you're showing them another path

[00:46:30] of forward and that will result in a win-win so don't keep casting letting people go and then

[00:46:38] you know going fishing again you know nurture the fish that you have. Yes,

[00:46:43] whilst we're using analogies I mean I was just thinking about it I mean I've been watching

[00:46:47] the Euro soccer championships for the last month and the English coach has had a lot of grief about

[00:46:53] playing players outside of their natural positions and as he's changed that and moved them across to

[00:47:02] be playing in there England have started playing better football and guess what we're in the final

[00:47:09] on Sunday maybe he's just going to let them play unfettered in their best suited positions

[00:47:14] and will win the whole thing you could see it the natural left-footed players playing on the right

[00:47:20] out of place they struggled it wasn't where they were used to playing in a team put them yeah put

[00:47:27] people in their best place. Yeah, going back to something that you said in your intro in terms

[00:47:32] of you you know you took this assessment when you were pointing something and and that sort of

[00:47:39] realigned you to you know this new career path so you really took to those results right you really

[00:47:46] trusted those results. So when we think about assessments today and the fact that ideally

[00:47:53] they are giving insight to the hiring team but hopefully giving you as the candidate insight

[00:48:00] as well or maybe the employee right yeah you're taking assessment internally but I guess I wonder

[00:48:06] over the course of your work are you seeing people use behavioral and psychometric assessments

[00:48:13] effectively when it comes to those you know the durability of those human skills because there's a

[00:48:20] lot of different ones out there and really any candidate technical or non-technical should be

[00:48:25] taking those you know the technical talent shouldn't only be taking you know coding

[00:48:29] assessments and technical tests but yeah I guess more generally just you know your take on the

[00:48:35] assessment space. The assessment space is really interesting to me this year I my prediction

[00:48:43] for the beginning of the year was the great recognition not the great resignation or the

[00:48:48] year of the great recognition which was assessing and recognizing the skills people had

[00:48:54] what they needed and their progress to close that hard or soft skill durable skills whatever

[00:49:04] not all assessments are created equal there are some good ones and there are some not so good ones

[00:49:11] there's a lot of work being done on can AI equal the research and the statistical

[00:49:19] reliability that some of these other psychometric tests have had to prove have developed for validity

[00:49:28] over the years and that's good but we're a long way from having the data to prove that an AI

[00:49:38] assessment of skill is as good as a pre-hire assessment that's been done to a thousand

[00:49:45] subjects and they had all of the kind of variables factored in and that there's a reliability fact

[00:49:52] for the pre-hire assessments that industry has had to prove for the last 30 years that yes this

[00:49:59] is a valid unbiased assessment and you can't sue me for using it right and making you take it as

[00:50:05] a candidate we do not have anybody yet who's publishing that data on their AI skills inferences

[00:50:15] of you're a level three on this Bob and we should otherwise we're letting them off the hook

[00:50:22] and if it was an appropriate standard before why is it not an appropriate standard today

[00:50:29] now and so the reliability and validity of those assessments I know some organizations

[00:50:34] are spending a lot of effort in the rigor of that data and psychological validation and accuracy

[00:50:45] but some are not someone just whipping it up with pre-built models out of the chat tpt store

[00:50:52] and think hey presto is a pre-hire assessment tool using AI I'm hearing candidates are using AI

[00:50:59] to actually complete assessments yeah ridiculous it's that's the state of the job market apparently

[00:51:06] that you have to have AI give your responses which should be coming from your human brain

[00:51:11] using AI anyway yeah and that's that's been on a very slippery slope for a while I mean

[00:51:17] Goldman Sachs had their infamous kind of battery of tests that somebody leaked out a kind of

[00:51:24] algorithm tool to help beat get through the Goldman Sachs tests brilliant investment bankers

[00:51:30] but that was probably three or four years ago there's probably at least and then there was the

[00:51:37] well I'm just gonna have my dad sit in on my interview and he can feed me the answers to

[00:51:41] give me a break now we're just getting that my AI agent will talk to yours and

[00:51:47] tell me when I should come in and you can make me an offer yeah there's all kinds of

[00:51:52] sneaky stuff going on in terms of the assessments and using AI and what have you I mean I

[00:51:59] obviously I've been spending a lot of time in responsible AI space and I've been learning

[00:52:04] a lot about the assessment space between my time with with Lumini and talking to folks like Charles

[00:52:10] Handler and others but the fact that from a leg legislation standpoint it's really about

[00:52:20] you know using AI or some algorithm to come up with some type of score or stack rank etc

[00:52:26] I don't know how you can say it's not part of the decision-making process because

[00:52:30] it clearly is but are they circumventing it by not coming up to the score

[00:52:37] probably it looks like a duck and it sounds like a duck it's probably a duck so yeah probably

[00:52:43] but I think I mean they're all full of the disclaimers of don't make your decision only on this

[00:52:50] data it's all those kinds of things also in terms of pushing off the liability but if you're

[00:52:56] going to make recommendations you've got to have some skin in the game you can't just say yeah

[00:53:02] go smoke this cigarette it won't hurt you what do you mean it's going to come back and bite

[00:53:08] you at some point so Gordon I think we uh I think we talked for longer than

[00:53:15] me anticipated my little counter here is definitely off but no it's obviously been a great engaging

[00:53:21] discussion I do have one final question for you of course because I ask it of all my guests but

[00:53:26] when you see the phrase elevate your AIQ what advice do you have for people to elevate theirs

[00:53:33] or what comes to mind when you hear that phrase yeah that's a great question I mean

[00:53:38] multiple sources is research if you only go find one place that's plagiarism right if

[00:53:44] you go to multiple places to find things that's that's learning and research so I think my

[00:53:51] comment is to go look around lots of conversations about AI it isn't all the doom and gloom there

[00:53:59] are some really interesting even just in your daily life thing it's not all about work I let's

[00:54:05] find something I mean I mentioned the BYO AI I mean and maybe that's where we're going to

[00:54:11] find real momentum as we'll bring our own tools of choice and our own AI to what it is we need

[00:54:20] whether it's planning a holiday or putting a power point together it's like skills start try it

[00:54:28] experiment I'm probably behind I'm behind the curve on many people but I'm ahead of it on a lot

[00:54:34] of others in terms of what I how I choose to engage and use it and find its frailties and

[00:54:42] argue with it about its inability to spell in graphics it's like am I really I mean that invested

[00:54:49] in this apparently excellent no that's a good advice they agree all right Gordon that is it for

[00:54:59] today thank you so much for joining me this has been great my pleasure it's been a great

[00:55:03] discussion excellent thanks everyone thanks for joining us today and we'll see you next time on

[00:55:09] another episode of Elevator AIQ thanks for life for now