The future of work is already here, and it’s moving fast, particularly in the U.S., according to Leslie Thomas from Kryterion. While Europe is focusing on regulating AI, the U.S. is more focused on building on the technology.
“We tend to really focus on innovation. If you look at some of the graphs that are out there, as far as the investment that the US has made compared to other countries, the U.S. is on top. So we’re just kind of running headfirst into it”, says Thomas.
In this episode, Kryterion’s Chief Psychometric Officer breaks down how AI is shifting the way we think about jobs, skills, and certification.
AI is expected to automate many routine tasks. A recent comment from Anthropic’s CEO claimed up to 50% of entry-level white-collar roles could vanish in five years. Large language models are already picking up tasks like document review, drafting emails, or data entry.
But that doesn’t mean people are being left behind. In fact, Leslie says many entry-level workers are already using these tools and adapting quickly. The real challenge? Helping those without the infrastructure or experience to catch up. Companies need to reskill and upskill their teams, not just chase new tech.
If you’re in a field that’s heavy on routine, your role will likely shift sooner. But Leslie offers a practical way to plan for it: look at your job through three buckets. What can only a human do? What could AI help with? What could be automated entirely? Then double down on the first two.
A tech job will change faster than one in healthcare. But across industries, change is coming. Kryterion is already seeing clients adjust their certifications to keep pace. They’re using AI to build better assessments, faster, and to reflect the skills employers need right now.
Those skills are not what you might think. The World Economic Forum lists the top ones as analytical thinking, creativity, flexibility, and resilience. Math and writing came in much lower. That shift matters. The people who thrive in tomorrow’s workplace won’t just know how to code, they’ll know how to question, adapt, and think critically with AI as a teammate, according to Thomas
She also sees AI helping solo entrepreneurs grow in ways they couldn’t before. Content, videos, pitches, like many are already doing. What once took weeks now takes hours. It’s opening doors for those who might be overwhelmed at the thought of tackling a business from all corners.
Certifications are evolving, too. Kryterion works with companies to define the minimum skills someone needs to do a job safely and well, whether that’s in tech, healthcare, or finance. And as agentic AI matures, those skills will keep shifting.
So, where does that leave you? If AI can draft your emails or sort your data, what does your job become? Are you learning how to work with AI, or waiting to see what happens next?
You can listen to the full episode on Spotify, Anchor, Apple Podcasts, Breaker,, Google Podcasts, Stitcher, Overcast, Listen Notes, PodBean, and Radio Public.
Find out more about Leslie Thomas here.
Reach out to today’s host, Erick Espinosa – [email protected]
Transcript:
Leslie Thomas:
My name is Leslie Thomas. I am the Chief Psychometric Officer at a company named Criterion. We’re a technology company that helps clients design, deliver, and administer high-stakes assessments. So specifically certification and licensure exams. If you’ve ever taken a driver’s license test, that’s an example of a licensure exam. We work a lot with tech companies on certifications as well.
Erick Espinosa:
Perfect. Leslie, thank you so much for joining me on this episode of Brains Byte Back to explore, like you mentioned, how AI is transforming the way we validate skills and testing in the workforce, which has changed a lot in recent years, especially with AI becoming more popular. Can I start by asking you how you ended up in this specific role? Because you’re the Chief Psychometric Officer, which is a term I haven’t heard before—could you explain a little bit about what that means?
Leslie Thomas:
Yeah, you definitely wouldn’t be the first that hadn’t heard of psychometrics. The joke is my family still really wonders what I do. So I should probably start with what psychometrics is. It’s actually a branch of psychology that studies the design, development, and evaluation of psychological tests that measure any type of psychological variable. That could be knowledge, skills, abilities, attitudes, personality.
And the interesting part is, even though a lot of people haven’t heard of psychometrics, if you’ve ever taken a college admissions test, a driver’s license test, a career interest inventory, personality test, or some sort of certification exam, your life has actually been touched by psychometrics. We’re the ones that help design and ensure that the scores are meaningful, accurate, reliable, and legally defensible.
Needless to say, I didn’t grow up thinking I wanted to be a psychometrician. Like most people, I had no idea what that was. I ended up going to graduate school for business psychology, also called industrial-organizational psychology. My advisor was doing research on the Myers-Briggs Type Indicator. For my master’s and dissertation work, I created a parallel version of the exam and also built an adaptive testing algorithm, which I later sold. That ended up being fairly lucrative.
After that, I worked at ACT—American College Testing. I eventually realized I didn’t enjoy large corporations, even though it was a great experience. From there, I started moving into more startup environments. I had my own consulting company, did pre-employment testing, performance improvement consulting, worked with Wall Street banks and large IT companies. Then I joined a startup in D.C. to build their testing and psychometric program, developing high-stakes exams. Eventually, I joined Criterion to build out that function there too.
Funny story—when I went to the D.C. company, they needed a title. It’s usually something like Chief Assessment Officer or Chief Science Officer. They were big Star Wars fans and wanted something with “CPO,” so we landed on Chief Psychometric Officer. It stuck. It’s a good conversation starter. I’ve always been entrepreneurial and enjoy working with companies to meet their business needs while also upholding quality standards.
Erick Espinosa:
And how has the role of AI impacted testing over the time you’ve been doing this?
Leslie Thomas:
It’s been pretty disruptive in two ways. First, it’s made cheating easier. We often compare it to playing whack-a-mole—you plug one hole, another pops up. AI has accelerated that. People can use browser plugins, deepfakes, large language models—there are all sorts of methods now.
On the other hand, AI has also helped improve how we monitor exams. We use AI-assisted proctoring that flags things for our human proctors. AI doesn’t get tired or distracted, so it’s been a big help for test security.
Another major shift is using large language models to help us create exams faster. As skills evolve faster—often because of AI—we need to update exams just as quickly. We need to validate new skills and get those exams to market in time.
Erick Espinosa:
It’s definitely a double-edged sword. You mentioned monitoring. From what I understand, the company does a mix of in-person and remote testing. I imagine AI would be especially useful for home testing?
Leslie Thomas:
Yes, exactly. We have a network of physical test centers for areas where people may not have strong internet or a distraction-free environment at home. But we were actually early adopters of remote proctoring, starting around 2006. It really took off during COVID when everyone had to shift online.
Remote testing allows people who work full-time or don’t live near a testing site to take exams from home. They use a webcam. Our proctors are located around the world and monitor the sessions live, with support from AI.
Erick Espinosa:
You do a lot of certifications. Is that becoming more popular now with AI taking off? Especially outside the traditional medical or legal fields?
Leslie Thomas:
Yes. At Criterion, we work a lot with tech companies. We were one of the first with a SaaS-based, self-service model. Tech companies with AI products are definitely driving demand.
The certifications are usually specific to their platforms—Google, Databricks, etc.—not general AI. We’re seeing traditional fields like healthcare and finance adopt AI too, but more slowly due to regulation. Those changes are coming, but at a slower pace.
Erick Espinosa:
What are some lesser-known or emerging careers you’ve seen take off because of AI? Any that you think will go mainstream?
Leslie Thomas:
A good example is prompt engineering. Most people hadn’t heard of it in early 2022. By the end of the year, it became a hot job—some people earning over $200,000. But now, it’s already changing as those skills get built into platforms.
That shows how fast things shift. So while AI ethicists, compliance auditors, and content curators are becoming more common, these roles will also keep evolving. The need to learn and adapt is more important than ever.
The World Economic Forum’s Future of Jobs 2025 report predicts 78 million net new jobs by 2030. Some jobs will go away, but new ones will take their place. Many of those haven’t even been defined yet.
Erick Espinosa:
It must be a strange time for both workers trying to keep up and students choosing what to study.
Leslie Thomas:
Absolutely. I’ve been doing research for an eBook and the key takeaway is basic AI literacy—just knowing how to use tools to get work done.
But what’s even more important are durable skills. Creativity, adaptability, learning how to learn. Those won’t go out of date like technical skills can. Social skills—leadership, collaboration, influence—will also be huge.
We’re going to see more cross-functional teams: tech, business, marketing, and finance working together, often using AI agents. That means more group work, and people need to know how to use AI responsibly and ethically. Human oversight, shared decision-making—all of that is becoming part of everyday work.
Erick Espinosa:
That idea of leadership at all levels is really interesting.
Leslie Thomas:
Yes, and that came up in my research too. AI levels the playing field. It gives more people access to powerful tools. That means employees at every level will need to know how to ask the right questions, guide the technology, evaluate results—things we usually expect from managers. With agentic AI growing, more people will need to manage, monitor, and govern these systems too.
Erick Espinosa:
That’s a great point. And since you work across global regions, do you notice any major differences in how AI is being adopted between, say, Europe and the U.S.?
Leslie Thomas:
Yes, it’s interesting. You know, a lot of it depends on, you know, with the AI. When we look at how it’s going to impact job roles, one of the big factors is, you know, where you’re located and how much regulation you have. So Europe passed the EU AI Act, I think it was last September.
It’s still, you know, they’re trying to figure out exactly how to interpret because then the standards are pretty general. And I think they’re looking at it again, because they don’t want in one sense, stifle innovation, but at the other end, they want to make sure that it works for humans, you know, that we’re careful in terms of how we implement it. And then you kind of go to the other end of it within the US and we, you know, don’t have that same regulatory framework.
You know, we tend to really focus on innovation. If you look at some of the graphs that are out there, as far as the investment that the US has made compared to other countries, I mean, it’s, you know, when you look at the scale, the US is like way up here, and then the other countries are down here. So we’re just kind of, you know, running head first into it.
And so, you know, I think within the US, we’re going to see, you know, a lot more impact on jobs, because we just don’t have as much of the regulations related to, you know, worker protections and things like that. So, you know, in this country, in particular, we’re going to have to figure out how to adapt, you know, and how to really focus on how do we employ it in a way that really puts workers first? How do we help people reskill and upskill in a way so that they feel comfortable with the change, just because there’s so much fear out there.
I mean, I think last week, as the CEO of Anthropic, Dario Amodi had mentioned that he thought within the next five years that AI was going to take over, I think, 50% of entry level white collar jobs. And, you know, of course, that made headlines everywhere. And it kind of makes sense, because the things that are more routine, or can be easily automated—data entry, reviewing, summarizing documents—those types of things are going to get automated. That’s what the technology does really well.
But I also think that people that are what we call entry level that are just entering the workforce, they’re also used to using these technologies. And so I think they’re going to be—they’re already using these technologies. And so I think that they’re going to also be able to adapt pretty readily.
So but it is something that, as a society, we really need to think about how we want to handle and manage, because some people don’t adapt as well to change. And that’s going to be a big factor. I think entry level jobs will always exist to some degree. It’s just now when you have, I guess, AI doing what we traditionally think is an entry level job, then people need to find new opportunities in that career. An entry level job is basically what opens the door for you to get a little bit of insight in what your colleagues are doing and testing it a little bit.
But if anything, I just feel like it won’t completely disappear. But it’ll just change in terms of what you’re being asked to do.
Erick Espinosa:
Right.
Leslie Thomas:
And I think in a lot of ways, I mean, the upside that can be a little difficult to think about just with all the fear mongering and stuff that’s going on is—a lot of the stuff that AI does, because it’s doing the routine task. I mean, a lot of times, that’s the stuff that people may not really enjoy doing.
And so, you know, it’s great at drafting emails or documents and things like that that might get you 50, 60, 70% of the way there. And then you get to react and refine that, instead of trying to come up with what I call the blank page problem, where you’re starting off from scratch with And so, I think it’s going to allow people to really focus on more high value tasks, and also things that are what humans are really good at—the creativity, the strategic thinking and those types of things. So I think that’s really important.
The other thing that I’ve noticed that I think is really interesting—and it’s amazing—is I’ve noticed a lot of people that don’t want to work in a traditional corporate environment. AI really allows your solopreneurs and your small businesses to really scale in a way that wasn’t possible just a few years ago.
So I have a friend that I was talking to over the weekend, that’s been trying to get a company off the ground for a couple years now, really been struggling—very smart, super motivated. And she was just saying it’s amazing what she’s been able to do with these large language models in terms of creating content, putting videos together, and doing all this sort of stuff that she said before would have taken her weeks and weeks to do.
So it’s really opening up a lot of opportunities for people. And I think that’s really—we’re gonna see, I think, a lot more people going out on their own. And whatever they’re really good at, whatever unique gifts or skill sets and talents they have, they’re going to use AI to really leverage and lean into that, which I think is really exciting.
Erick Espinosa:
I kind of want to make a little turn here back into the certification process, because I’m kind of interested. I guess when you meet with these companies, and you’re building these certification processes, what are some of the main things that you try to point out? Right? Because this is very new to a lot of companies. They come in, they’re like, “I want to develop something new.” How does that process look from the beginning?
Leslie Thomas:
Yeah, so one of the things with certification—and even with licensure—that a lot of people may not originally understand is that one of the things we’re trying to do in different job roles is determine: What is the minimum standard of competence? And we’ll say “minimal competence,” and sometimes it has a poor connotation because we say “minimal,” but it could be a very high standard of performance that’s being required.
What certification and licensure does is, it allows an organization—whether it’s an industry association for nurses or doctors, or if it’s a tech company—to figure out: What’s the standard of performance that we’re going to require of all people that are certified, that we’re going to require in order for them to implement our technology, or for them to get preferential treatment as an administrator of our technology, or something like that, or to be able to practice medicine, or work in a certain healthcare or finance area.
So one of the things is really sitting down with them and having them define: What is the pain point that you’re trying to solve with the certification? What’s the business problem? Because ultimately, it is a business, and oftentimes also for public welfare and safety. And then working with their subject matter experts to define what is that minimal level of competence they want to ensure people have in order to become certified or licensed in that particular area.
So it’s a little different from some testing that’s out there where you’re trying to predict job performance or who’s going to be the best at something. This is really setting up a standard of performance and then trying to make sure that people who meet or exceed that standard are qualified and minimally competent to perform that particular job function or activity.
Erick Espinosa:
Do you see any trends—because you mentioned agentic AI—do you see any trends shaping the certification process, like in testing in the next three to five years?
Leslie Thomas:
Yeah, it’s hard to predict with agentic AI. My team is working on incorporating large language models into our product to enhance test development and those types of things, and we’re already seeing value in using the agentic piece of it to help us create test content.
So rather than trying to put everything into, you know, one prompt or whatever to create a number of agents to do it and so forth. So, we’re starting to incorporate it. It’s difficult to say how it’s going to impact businesses because a lot of this right now, we’re kind of going through this awkward transition period, I think, from a technology perspective where, with the tech companies that I work with, the technology is getting very advanced very quickly, but then the businesses are kind of catching up.
Maybe they don’t have the data or the infrastructure or, moreover, even the people with the skill sets to help them adapt and adopt AI. And so, it’s difficult to know what that timeline is going to be—what that’s going to be like.I think that for those areas like in tech that are at the leading edge, it’s going to impact those certifications more quickly than it will be, say, in healthcare for something like a surgeon or something, which is more regulated and more slow to change.
I think that’s where we’re going to see it impact those types of job roles. It’s difficult to say exactly how it’s going to impact the job roles because it kind of depends on what the agents are really good at and then how quickly companies start adopting those particular processes or agents.
Because companies that develop the certifications want to certify on what the companies are doing at that point—or get them ready for the future, the near future, so to speak, if that makes sense.
Erick Espinosa:
Definitely. There’s a lot of variables. I still feel like we’re in the very early stages of especially what’s happening. Like, obviously, there’s a technological revolution that’s basically happening right now, but we’re seeing a lot of—even when it comes to AI—like what the future holds, but it seems to be coming, like, very fast.
Leslie Thomas:
Yes. Yeah, it’s difficult. I mean, that was one of the things when I did a lot of the research on the eBook. There are certain factors that we can—we talked a little bit about them here.
It’s just: Do you live in a country that’s highly regulated with AI or not? What’s the level of investment? What type of job are you in? If you’re in the tech industry, then your job role is going to change a lot quicker because they’re the ones that are really embracing it—or if you have a job that has a lot of routine tasks.
In fact, I tell people that it’s helpful to go through kind of what I call the three-bucket exercise, where you look at those tasks in your job that are human-only that probably will stay that way for a while; those tasks that can be supplemented or augmented with AI, where you can use a co-pilot or something else to help you do that part of your job better; and then those things that are either automated or could be automated in the near future.
And then really lean into those first two buckets—really focus on those—because that’s the area, I think, when the transition or when job roles start transitioning, it’s likely going to be more gradual than some of what we’re being led to believe. And it’s going to be one of these things where, you know, some part of your job gets automated. Then the AI evolves, and this AI can help you do this part of your job better. It can help you with decision-making, content, all those types of things. And then this area is still left for human-only.
In fact, the World Economic Forum—and that report I referenced earlier, The Future of Jobs 2025—they predict by 2030 that our jobs will basically be a third, a third, and a third. So a third will be human-only, a third will be augmented with AI (where it’s human and AI), and then a third will be automated.
And they’re looking across a lot of job roles. So again, any particular industry or job role, those percentages will be different, but that’s just kind of an overview of where they think things are moving.
Erick Espinosa:
I think it’s a scaling for efficiency, basically. So you’re just trying to make your jobs a little bit more efficient and focus your energy on things that need the human mind rather than the AI aspect to it.
Leslie, what was the name of the book that you mentioned—your eBook?
Leslie Thomas:
Yeah, it’s AI and the Future of Work: An Assessment. So I did a bunch of research on just the work side of it and then also looking at how that’s going to impact how we assess those job skills.
I have a chapter in there on what are the skills from an AI perspective that’ll be really important. And really, for our jobs, a lot of it’s going to focus on some of those durable skills.
And in fact, going back again to the World Economic Forum—their top six skills—the top five are actually these more what we call “durable” or sometimes “soft” skills: analytical thinking, creativity, resilience and adaptability, flexibility. Interestingly enough, out of 26, I think 21 was math, reading, and writing, which I just… yeah, I know, it was pretty surprising.
Anyway, I did a lot of research on how it’s going to impact—and also what are those types of skills that people really want to lean into. A lot of it is, as I said, some of the leadership at all levels, some of the durable skills I go through, and then also the technical literacy or understanding AI technology.
Specifically for what your job role is. Because I think more and more, it’s not going to be so much people’s traditional technical skills that’ll be important in a job—it’ll be their ability to use AI technology to be more productive and efficient.
And so it’s going to be understanding which tools to use when, what are the risks, benefits, limitations of those tools. And then the analytical or critical thinking is going to be extremely important.
Because as you’ve worked with large language models, I’m sure it’s really about critically looking at the output and figuring out: Does it make sense? How can it be improved? How can I ask a question to make it better—and iterating? So yeah, AI and the Future of Work: An Assessment.
Erick Espinosa:
No, it sounds like great, valuable insight. I appreciate it.
Leslie, thank you so much for joining us. If somebody is interested in learning more about this certification process—maybe wants to pick your brain—what’s the best way of reaching out?
Leslie Thomas:
Sure, you can reach me at my email: [email protected]. That’s K-R-Y-T-E-R-I-O-N online.com. And you can also go to our website, kryterion.com, and look us up there.
Erick Espinosa: And I’ll make sure to include that link at the bottom below. Leslie, thanks again.
Leslie Thomas: Thank you. It was really a pleasure to be here.
Disclosure: This article mentions a client of an Espacio portfolio company.
