00:00Meanwhile, the AI economy, well, it's moving fast, and now we may have a clearer way to track how it's
00:05changing the way people work.
00:07ADP has partnered with the Stanford Digital Economy Lab to launch the Canary's Dashboard.
00:12It's a real-time indicator designed to show how AI is reshaping different occupations based on actual labor market data.
00:20Joining us for more is one of the researchers involved in this project, Neela Richardson, chief economist at ADP,
00:25former senior economist, actually at Bloomberg, and Neela, present. You are monitoring the present, not the past.
00:32But what data do you take in? How are you monitoring this?
00:35Well, first of all, we are thrilled because AI is set to be one of the most consequential technologies the
00:42world has ever seen.
00:44And yet, we have very limited ways of measuring impact.
00:48And so, this is why I'm so excited to partner with Stanford Digital Economy Lab, led by Eric Bern-Jolson,
00:55on these AI indicators,
00:57and namely the Canary's Dashboard, which tracks the impact of AI on occupations in almost real time.
01:04This is about moving the conversation about AI's impact from what we think to what we know and what we
01:11can measure with the data.
01:13Neela, there was a section of the labor market that just jumped off the screen at me, and that is
01:19the early career workers.
01:21These are people aged 22 to 25.
01:23And there is a, I don't know how you would put it, a bifurcation.
01:26Right.
01:27Right. Industries that have exposure to AI, industries that have nothing to do with it whatsoever.
01:31What is the data telling us there?
01:33And if you can, the why?
01:36The why?
01:37Well, the important part of this data series is that it is able to categorize over 700 occupations by AI
01:44exposure in a very granular way.
01:47But it's more than that, because it's not just how the AI is affecting occupations, but it's about how AI
01:54is affecting workers, people, young careers.
01:58So, we're able to use the demographics in the ADP payroll data and segment it by early career, 22 to
02:0426.
02:04And look, AI, you know this as well as anyone, has two different roles in a business context.
02:11It can augment.
02:12That's the old school story.
02:14That's the dinosaur story of technology.
02:16How do you, sorry, automate work?
02:19That's the old school story.
02:21The new school story, the frontier, is how to augment.
02:24And so, what AI does for early career in AI exposed fields, it looks like it's automating certain tasks.
02:30So, the trick is, how do we move from automation to augmentation and look for those higher value tasks, higher
02:39value work?
02:40And the result is that in that space, employment is contracting.
02:44Right.
02:44So, for AI exposed careers, there is a contraction for early career in, like, software developers.
02:50So, let's take software developers.
02:53Since the rollout of ChatGPT in November 2022, this dashboard shows there's been a 20% decline in early career
03:01software developers.
03:03But when you look at older workers, no decline at all.
03:06In fact, you're seeing growth.
03:08That shows you that there is a disparate impact here.
03:11For skills and tasks that are easily automated, you're seeing an effect.
03:15That's the early career.
03:17But for work where it's more complex, AI becomes a helpmate, a co-worker, an augmentation tool, as opposed to
03:24an automation tool.
03:26Look, this is going to be released every Wednesday after Jobs Week, so it's monthly data.
03:30How are you thinking about when you realize an industry is becoming AI exposed?
03:35At the moment, you've been so fascinated with the fact that we've got developers, customer services.
03:39But we're trying to understand, sort of, where's the next?
03:42How are you seeing low AI exposed operations starting to change or become AI exposed?
03:47That's a great question.
03:49And that's really the purpose and mission of this work.
03:51It is to track value creation in real time.
03:54And the thing about it is you can't track it in a macro way.
03:57You can't track it in the markets.
03:59It's kind of go to all of this, yeah.
04:00Yeah.
04:01Market IPO creation, value creation is very different to how it affects the real economy and what people are really
04:07experiencing at work.
04:08And so, at the task level is where you see value creation.
04:12And you see that in certain complex jobs.
04:15So, let's move on from software developers, maybe look at radiologists, where AI becomes a really important diagnostic tool.
04:25And you can see that value creation and delivery, helping them concentrate on the work that is necessary for human
04:32-to-human interaction,
04:33as opposed to simply diagnosing different patterns, which AI is good at.
04:38So, the key for employers is how to extend human capability, not limit it, not replace it, but extend that
04:46capability to new tasks and new value creation.
04:49We have a good case study for that.
04:51Bloomberg's Remain Bostic just spoke to the IBM CEO about this exact point.
04:56Let's listen to what he had to say, and then you can say if it shows up in the data.
05:00Using AI tools now, we can probably add 10 points of profit synergy on day one, because the amount of
05:07time it used to take to move contracts over,
05:10to do all of the sales automation, to do all of the revenue forecasting, all of that now, using AI,
05:17can be shrunk down to literally a few weeks.
05:20When I listen to that, I just can't draw a conclusion.
05:23Are we talking about role elimination?
05:26Are we talking about a boost of productivity, which Caroline and I were told by SFF President Mary Daly last
05:32week isn't showing up in the data yet?
05:34How do you read the corporate speak in this job market?
05:39The way I read it is this.
05:42AI has the ability to reshape work, and yet it's still a tool.
05:47That decision lies with the employer.
05:51And so this data is about empowering employers to make that decision.
05:56Is AI going to be your efficiency tool?
05:59There's a clear case for that.
06:01There's a clear use case that AI is making certain work more efficient.
06:05You can do more with less.
06:07But I think AI has the potential for more than that.
06:10It is a productivity tool in the sense that it enhances work.
06:14It makes work better.
06:16It makes problems easier to solve, and therefore you can tackle more problems.
06:21If it's an augmentation tool, that is something completely different.
06:26And hopefully data will help employers find that value creation within their own businesses.
06:31So right now, I think the narrative is really, really wide.
06:35The data can help anchor it on the truth of the moment.
06:38All right.
06:38All right.
Comments