Why do promising AI initiatives fail even when the technology works? Behavioral scientist Dr. Gleb Tsipursky explains why employee fear, professional identity, shadow AI, and poor leadership decisions can quietly sabotage AI adoption inside a company.
In this episode of The Mason Duchatschek Show, Mason talks with Dr. Tsipursky, CEO of Disaster Avoidance Experts, author of seven books, and author of the forthcoming The Psychology of AI Adoption at Work: From Resistance to Results, published by Georgetown University Press.
Dr. Tsipursky spent more than 15 years in academia as a behavioral scientist, including faculty appointments at The Ohio State University and UNC-Chapel Hill. His work and insights have appeared in Harvard Business Review, Fortune, and Forbes, and The New York Times. He has also advised organizations on decision-making, workplace change, artificial intelligence adoption, and the future of work.
In this conversation, Dr. Tsipursky makes a case business leaders need to hear:
The biggest obstacle to AI adoption may not be the technology. It may be human psychology.
Employees may fear they are being asked to train their own replacements.
Others may feel that AI threatens the very skills that define their professional value.
Still others may already be using AI successfully but hiding it from coworkers and management, creating what is increasingly known as shadow AI.
For CEOs, business owners, HR leaders, executives, sales managers, and team leaders, those psychological barriers can mean the difference between buying AI tools and actually getting measurable value from them.
In this episode, you’ll learn:
• The 3 major psychological barriers that can derail AI adoption
• Why employee resistance to AI is different from resistance to previous workplace technology
• The mistake CEOs make when they buy AI tools and simply tell employees to use them
• Why fear of job loss can undermine AI implementation
• How AI can threaten an employee’s professional identity
• What “shadow AI” is and why leaders should pay attention to it
• Why some employees hide successful AI usage
• How leaders can reduce resistance without dismissing legitimate employee concerns
• Why starting with employees’ most frustrating tasks can accelerate AI adoption
• How prompt libraries can make AI easier to use across teams
• How AI agents may reshape sales, meeting preparation, performance management, and business workflows
• Why future managers may need to lead people who manage AI agents
• How prompt engineering can improve communication with humans as well as machines
Connect with Mason Duchatschek
Website: https://masonduchatschek.com/
LinkedIn: https://www.linkedin.com/in/masonduchatschek/
Workforce Alchemy Magazine
Subscribe here:
https://workforcealchemymagazine.com/
The Mason Duchatschek Show Podcast
Listen and subscribe:
https://shows.acast.com/themasonduchatschekshow
#AIAdoption #ArtificialIntelligence #AILeadership
In this episode of The Mason Duchatschek Show, Mason talks with Dr. Tsipursky, CEO of Disaster Avoidance Experts, author of seven books, and author of the forthcoming The Psychology of AI Adoption at Work: From Resistance to Results, published by Georgetown University Press.
Dr. Tsipursky spent more than 15 years in academia as a behavioral scientist, including faculty appointments at The Ohio State University and UNC-Chapel Hill. His work and insights have appeared in Harvard Business Review, Fortune, and Forbes, and The New York Times. He has also advised organizations on decision-making, workplace change, artificial intelligence adoption, and the future of work.
In this conversation, Dr. Tsipursky makes a case business leaders need to hear:
The biggest obstacle to AI adoption may not be the technology. It may be human psychology.
Employees may fear they are being asked to train their own replacements.
Others may feel that AI threatens the very skills that define their professional value.
Still others may already be using AI successfully but hiding it from coworkers and management, creating what is increasingly known as shadow AI.
For CEOs, business owners, HR leaders, executives, sales managers, and team leaders, those psychological barriers can mean the difference between buying AI tools and actually getting measurable value from them.
In this episode, you’ll learn:
• The 3 major psychological barriers that can derail AI adoption
• Why employee resistance to AI is different from resistance to previous workplace technology
• The mistake CEOs make when they buy AI tools and simply tell employees to use them
• Why fear of job loss can undermine AI implementation
• How AI can threaten an employee’s professional identity
• What “shadow AI” is and why leaders should pay attention to it
• Why some employees hide successful AI usage
• How leaders can reduce resistance without dismissing legitimate employee concerns
• Why starting with employees’ most frustrating tasks can accelerate AI adoption
• How prompt libraries can make AI easier to use across teams
• How AI agents may reshape sales, meeting preparation, performance management, and business workflows
• Why future managers may need to lead people who manage AI agents
• How prompt engineering can improve communication with humans as well as machines
Connect with Mason Duchatschek
Website: https://masonduchatschek.com/
LinkedIn: https://www.linkedin.com/in/masonduchatschek/
Workforce Alchemy Magazine
Subscribe here:
https://workforcealchemymagazine.com/
The Mason Duchatschek Show Podcast
Listen and subscribe:
https://shows.acast.com/themasonduchatschekshow
#AIAdoption #ArtificialIntelligence #AILeadership
Category
🛠️
LifestyleTranscript
00:05Welcome to the Mason Dukacek Show, presented by WorkforceAlchemy.com, where business leaders
00:12discover practical ways to select, engage, and retain their best employees.
00:16Today's guest is Dr. Gleb Sapersky, a behavioral scientist and CEO of Disaster Avoidance Experts.
00:25He's also the author of seven books, including the forthcoming, The Psychology of AI Adoption
00:31at Work.
00:31His insights have appeared in Harvard Business Review, Fortune, Forbes, and the New York Times.
00:38He helps leaders understand why AI adoption succeeds or fails inside of real organizations.
00:45Dr. Sapersky, welcome to the show.
00:48Thank you so much for having me on, Mason.
00:50I appreciate it.
00:51It's a pleasure.
00:51So, you argue that many AI initiatives fail for human reasons rather than technical ones.
00:58What is the most surprising example that you've seen of a promising AI project being undermined
01:06by employee behavior?
01:07What happens is that people are given an AI tool that's specifically customized to a workflow,
01:15and they're supposed to check the output of that workflow.
01:20So, for example, one thing I've seen is with emails to prospects.
01:26So, it's an example where the AI tool looks for an investment management company, wealth
01:34management company.
01:35It looks for various triggers.
01:39For example, an announcement of a business acquisition or an announcement of a divorce.
01:43And this is an opportunity for an investment management company to come in and offer its services.
01:51With a business acquisition, the former owner would just come into a lot of money,
01:57and he or she would then need to manage it effectively.
02:02So, that's an opportunity for an investment management company or with a divorce.
02:07One of the partners would usually be managing the money, and then the other partner would
02:12now need to learn how to manage the money for a significant sum of money.
02:17And this is, again, an opportunity for an investment management company to come in and do its
02:22thing of managing.
02:24And so, I worked with the company to develop a tool that would research various events and
02:32would draft an automated email based on the event that occurred to send to the prospect.
02:40They would find the prospect's email address, they would draft the email, and then the account
02:46manager is supposed to evaluate it and send it and tweak it, improve it in the process.
02:54But what happened, and this was before it came in, is that the account managers would just
02:58automatically send the email.
03:00And so, the email was reasonably customized, it was well done, but it wasn't in their voice.
03:06And so, as a result, once people started responding and they started talking to the account manager
03:14by email, they were confused because the initial email was not like how the account manager
03:20sounded on later emails.
03:21And so, that caused a problem with client retention.
03:26It's like, why are you sounding different?
03:29And so, that's a problematic tool because it's not customized.
03:34The account managers weren't trained on how to use the tool effectively.
03:40They weren't trained on how to change it to sound like them because each person sounds different.
03:45More broadly, the underlying issue, going back to the psychology, was that people were reluctant
03:52to adopt this tool.
03:54They were hesitant.
03:56They felt like this tool was a threat to their professional identity.
04:01They felt proud of being able to prospect for new people, to reach out to them.
04:07And they felt that, hey, this tool is going to take my job and it's going to threaten who
04:13I am as an employee.
04:15And so, you have these psychological tendencies that made people reluctant to even get training
04:23on the tool.
04:24So, they were given this tool to use.
04:26And this tool is a really good tool to use, obviously, if you know how to use it well.
04:32But because of these psychological blockers, they were not using it effectively.
04:38And the psychological blockers, there are three big psychological blockers with AI adoption.
04:44And they're not the typical psychological blockers that people think when you're thinking about
04:48change management and the resistance to change.
04:50Because a typical psychological blocker with using new technology is that it's a hassle to
04:57learn.
04:57When you're adopting a new CRM, you need to learn about a number of places to click, a number
05:02of places to figure out how to do things differently.
05:07That might have been part of the reason why that last initiative you just described me failed.
05:11I could see multiple things going wrong with what you just described.
05:15A situation where some people might have over-relied on the AI and said, it's AI, it'll do it great
05:20and really embraced it and let it do its thing and not having their human voice and having the
05:26disconnect as a result.
05:29So, I could see why some people might just be over-relying on it.
05:32And I could also see where some would be hesitant like, hey, this is going to take my job.
05:36And they're not wrong to feel that way.
05:39When these companies are out there saying, oh, AI is not going to take our job, it's
05:42going to enhance it, justifiably so, there's a lot of employees that are going to call bull
05:46on that and say, no, this is going to take my job.
05:49But in this particular example, this is one of those exact examples where if they were
05:54over-reliant on AI and didn't have humans to go in and look at what was being said in
06:00their voice to their customers and having that mismatch, then obviously it was costing
06:05them to not have the human.
06:06And so, if I were an executive at that company, I would use that as an example of, yes, I
06:13acknowledge that your fear of change is real and that this could be a fear of losing your
06:17job.
06:18But if machines could do all this, well, you've let it try.
06:23And this was a disconnect with our customers and we lost business because of it.
06:28And we're not in the business of losing more business.
06:32So, let's help you get better at using this and find ways to work together to use artificial
06:37intelligence with real intelligence in a way that helps everybody.
06:41But that's going to take time because we're kind of in that transition where people are
06:45still trying to figure out what does this mean for me and what does this mean for the
06:48customer?
06:49How do you know when resistance to AI is legitimate warning sign and when it's just simply fear
06:55of change?
06:55So, let's talk a little bit about that fear of change.
06:58And so, the point I think it's important to realize is that fear of change previously
07:03did not have to do with fear of losing your job.
07:06And so, that's kind of fundamental.
07:08When you look, use a CRM, when you're adopting a new database management tool, when you're adopting
07:13a new CRM, an ERP, HRM system, any of these systems, you are not going to lose your job
07:19because you adopt that system.
07:21You're not going to lose your job because you've transitioned from using various Excel
07:24spreadsheets to look up customer information to using customer relationship management
07:28software to look up and log customer information.
07:32And so, when HR personnel, and those are the kind of things that people adopted previously.
07:39So, when HR personnel, when organizational change management professionals, when leaders
07:43in operations roles, when they think about change management resistance, they think about
07:48it's difficult to learn and people feel it's a lot of hassle and they don't want to let
07:52go of their current ways of working because it's a trouble to learn this new way of working.
07:59Now, that is what they're used to.
08:01That's their mindset.
08:02That's how they're thinking.
08:04This AI system is completely different.
08:07It's really easy to learn how to use AI tools.
08:09So, it's much, much easier to learn once you have the mindset shifts.
08:12Once you understand, you can talk to the AI tool.
08:14You can ask it how to do things.
08:16You can change it and evaluate it and it's super easy to learn once you get comfortable
08:21with it.
08:22But the key issue is the emotional blocker.
08:25And there are three emotional blockers.
08:27So, one of these is fear and anxiety.
08:29Fear and anxiety around job loss in a way that there wasn't around previous technology.
08:34This is a new thing.
08:36Leaders aren't thinking about it.
08:37HR professionals aren't thinking about it sufficiently.
08:39So, that's one of the groups.
08:42In my new book, The Psychology of AI Adoption and Work, From Resistance to Results, I studied
08:47this in over 100 consulting projects, including over 600 focus groups and many thousands of
08:52survey responses.
08:53I identified a psychographic profile.
08:55I talk about that in Chapter 6.
08:56It's called the AI Alarmists, for whom this is the major concern.
09:01So, that's one.
09:02Now, the second one is going to be people who feel professionally threatened by it.
09:08They're not so afraid of losing their job.
09:09But looking up information in a CRM isn't a threat to who they are as a professional.
09:15But the ability to write an effective email, the ability to make an effective sale, the
09:21ability to create a great image, to write a good article, to write a great report and
09:27do a great financial model, those are threats to who they are as a professional.
09:31And they don't want to let go of those tasks.
09:33And that is the second big obstacle.
09:36Let me see.
09:37And the third big obstacle is not those two obstacles.
09:41That's going to be the pragmatic resistors for people who are looking at Chapter 6 in
09:45my book.
09:46And the third big obstacle is going to be not to individual adoptions.
09:50The third big obstacle is for team adoption.
09:52And that obstacle has to do with the emotion of shame.
09:56People are ashamed of acknowledging that they used AI to draft their emails.
10:01They're ashamed of talking about how they used AI to do great marketing cut, to do great
10:07sales cut, to write great articles, great reports, images, and so on.
10:11And so, as a result, you have what's called shadow AI usage.
10:16So people are using AI, but they don't want to talk about it.
10:20And so, as a result, you have drop-offs at the team productivity level.
10:25People are individually more productive.
10:27But then, as they get done with their tasks, they just sit around and go and check out Facebook
10:33or something like that, because they don't want to talk to their supervisor about how
10:37they're using AI.
10:37They don't want to talk to their team members.
10:39So they get the same, they're expected to complete the same workload.
10:42And they're using AI to make it much more faster.
10:45And so you have a loss of team and organizational productivity while you have increased individual
10:51productivity.
10:52And so that shame, that social stigma of using AI is a third big emotional blocker.
10:57And that all leads to what the new study from MIT showed, that 95% of AI pilots don't show
11:04return on investment compared to the kind of resources invested into the AI tool.
11:10So that's a big, big thing to think about.
11:12So leaders, job professionals, folks of all sorts need to think about those three emotional
11:17blockers.
11:18And this may be a one-off.
11:19I'm curious what your take on this is.
11:21But I know one particular business owner who really, for lack of better term, kind of
11:29demeaned one of his IT professionals.
11:32It's like, can't you just have AI do this?
11:35I could do this with AI, or I could just do that with AI, implying that the IT guy should
11:42be able to do all these amazing things in languages he hasn't learned yet, which he probably
11:47eventually could figure out with help of AI at some point.
11:49But the demeaning way in which the owner talked to this IT professional, like, oh, you should
11:55just be able to do this.
11:56You should just be able to do this.
11:57AI can do this.
11:57AI can do that.
11:59And he's like, hey, I can tell you how to ride a bike too, but that doesn't mean you're
12:02not going to crash.
12:04Yeah.
12:04And there's a big difference between being told what to do and being able to do it.
12:09I could do a whole bunch of work on AI that say, hey, how do I play basketball like
12:13Michael Jordan?
12:14But is that going to make me play basketball like Michael Jordan?
12:17Absolutely not.
12:18And I just thought that was a very interesting conversation between an owner who just automatically
12:25assumed that AI was so much more valuable than his real employees.
12:28Real people still have to execute these things in real ways, whether it's riding a bike or
12:33playing basketball or programming, fixing computer programming problems in different
12:38languages.
12:39You still have to have a human to execute it.
12:41And I hope for those that are listening that may not run a business but are involved in
12:46leading or managing it, that they remember that and that it helps eliminate unnecessary
12:53friction between them and the people that are executing the stuff and may be fearful of
12:59job loss or feel threatened.
13:01What is the biggest mistake CEOs make when they announce an AI initiative to their organization,
13:06in your opinion?
13:07Yeah.
13:07So I think what you pointed to is really important.
13:12What I see in terms of executives making a mistake is saying, here, I just got you all
13:19Claude licenses and Claude is the most popular tool right now.
13:21Go ahead and use AI.
13:24That's the biggest mistake I see.
13:25And the executive then hopes that everyone will go and jump on it and use it the way she
13:29or he had.
13:30Because the executive is excited about it.
13:33They see they will be in the psychographic profile of efficiency seekers in that chapter
13:40six.
13:40They are excited about it.
13:41They're enthusiastic.
13:43They think that, hey, there's so much that could be done.
13:46And they're not thinking from the perspective of their employees.
13:48They're not thinking about the job question.
13:51A number of people are thinking about, oh, you want me to use an AI tool to train a replacement
13:57for myself?
13:57Or from the perspective of the, that's the AI alarmist or the pragmatic resistor.
14:03Oh, I am proud of doing all of these tasks.
14:06You want me to do tasks that I'm proud of?
14:08Like, what am I?
14:09And so thinking about how people are not using AI tools and just the way that AI tools are
14:16announced and told, everyone told to use it.
14:18That's the biggest mistake I see.
14:20So what should a CEO say when employees believe that AI adoption is really a plan to reduce
14:25headcount?
14:25There are two things that need to happen to address the AI alarmist, like a graphic profile.
14:31And so I talk about that in chapter 789.
14:34Now, what you want to do is, if possible, and most companies can do this, make a commitment
14:39to not fire anyone because AI is making your company more efficient, and instead focus
14:44some growth, we have a recent study coming out showing that AI companies, that companies
14:50that adopt AI quickly and effectively are actually growing at 10% faster in terms of headcount
14:56compared to others in their industry that aren't adopting AI effectively.
15:00Now, the interesting thing is they are growing faster compared to how much revenue.
15:07So they're growing, their revenue and their top line and their profit, their bottom line
15:12is growing faster than 10% because they're able to use less people to get more things done.
15:19But overall, they're growing compared to their competitors, and they're thus able to hire
15:26more people to do more things.
15:28So 10% faster.
15:29So it's an accurate statement.
15:30If you're able to make it, most companies are in an industry where they can successfully
15:34add competitors.
15:35There are some companies where there's a very limited market for some reason, and they're
15:40not able to make that commitment.
15:41But most companies are able to make that commitment to not focus on cost-cutting, to focus on growth.
15:46That's one.
15:48The other element here that's really important is to flip the script for the AI alarmists.
15:53Currently, they're thinking that, oh, if I train this tool, then I will be replaced.
15:58But what they need to realize is that the future belongs to those who use AI tools effectively.
16:03The future is going to be, all of us are going to be managers of agents, AI agents, right
16:08and left.
16:08And we can talk about that in more depth.
16:10But basically, what you need to do is tell them that your job will be safe as long as
16:15you learn how to use AI effectively.
16:17No question about it.
16:18Your job will be safe.
16:19You learn how to use AI tools effectively.
16:21You'll be more productive.
16:22You'll be more effective.
16:23Your job will truly be safe.
16:25And that's regardless of whether you're in a market where it's impossible to grow because
16:30and you need to focus on efficiency and cost-cutting because the people who will be let go are those
16:35who are not using AI efficiently.
16:37And even if something...
16:39Go ahead.
16:40It makes me think of...
16:42I mean, it's kind of...
16:43And I understand the fear now.
16:45Like, I get it.
16:46And I understand the optimism.
16:48I get that too.
16:50And I really think that you were hitting on something earlier.
16:54And I want to go back to that because I think it's very, very important.
16:57And I don't want to just zip over it.
16:59But if you look back, I can remember when the internet, email was new.
17:05Can you imagine someone say, oh, I don't want to learn how to do email because that could
17:09be a threat to my job.
17:11Or like you said, when it went from people using spreadsheets to CRM systems.
17:17Oh, I don't want to use the CRM system because that's a threat to my job.
17:21Or for that matter, if you want to go back further and say, why would I use computers?
17:26Like, I don't want that to be a threat to my job.
17:28And we can look back in hindsight and laugh at comments like that because exactly what
17:35you're saying is like, this is the way the future is going.
17:39And these tools and these technologies are just going to shift the way things are done
17:44so that you can do more, better, and faster.
17:48And if you resist it, then it's literally the equivalent of going back in time and say,
17:52oh, I don't want to learn computers.
17:53Or, oh, I don't want to learn how to do email.
17:56Or I don't want to know, I don't even want to know how to do spreadsheets.
17:59And when you look at it that way, it's almost laughable.
18:02So I really think that you did a very nice job of explaining the transition to people.
18:08And it's like, look, this is just the newest evolution of work.
18:13Could you maybe share a story where a leader changed the way AI was introduced and turned
18:18resistance into genuine participation and enthusiasm even?
18:22Sure.
18:22So I worked with a number of companies where the get close for everyone and resistance is
18:29the result.
18:30If possible at all, and large majority of companies can do this, focus on growth, not job
18:34cutting.
18:34And then talk frankly about your future is going to be secure if you learn how to use AI tools
18:39effectively.
18:40So that's going to help address most of the concerns of that AI alarmist.
18:45You have the carrot and stick.
18:46Now, you also need to address the concerns of the pragmatic resistors.
18:50And so how leaders can do this, and what I've worked with leaders in doing, is to not focus
18:56on automating the kind of activities that people feel proud about, that represents part of their
19:02professional identity, that represents part of their creative identity.
19:05Instead of saying, how can we effectively use AI to accomplish these really important tasks
19:10for you, writing articles and reports and so on, do a survey that says, what are your
19:16most annoying three tasks?
19:18What are the things that you do where you have to look up across a whole variety of systems
19:23and gather data together across a variety of systems and pull it together in order to
19:28do the important tasks that you want to be doing?
19:30What are the activities that you would like to get off your plate?
19:34So those sorts of questions.
19:36You want to do a survey to understand those sorts of questions.
19:39And then you want to create a prompt library.
19:42So a set of prompts that would help people accomplish these workflows that are going to be skeleton
19:49prompts, meaning ones that people can customize to their needs that are going to accomplish these
19:54specific workflows that people feel annoyed by, frustrated by, that they want to get off their
19:59plate because AI is a super flexible tool.
20:02You can get it to do almost anything that you, in white collar work, that doesn't have to do with
20:08the physical world.
20:09You can get it to do almost anything.
20:10And if you help people automate the tasks that are super annoying to them, frustrating, that they
20:17want to get off their plate, that's where you'll get the pragmatic resistors to start actually
20:22using AI tools.
20:24Everyone has aspects of their job that they hate.
20:26You can get AI to do most of these things.
20:29So what you can do is focus on those, then train people on those specific activities, give them
20:35prompt libraries, train them on those, and then people will start getting into it.
20:39Over time, they will naturally expand to other areas, things that they feel are closer to their
20:45central of who they are, because they feel how efficient the AI tools are.
20:50They're doing this.
20:50Well, okay, of course, I'm going to do it to those things.
20:53But don't start with it.
20:55If you start with it, you get a lot of resistance.
20:56Okay.
20:57So before we continue, I want to share an opportunity for listeners who are ready to
21:02put better workforce practices into action.
21:05They can apply for the Workforce Alchemy Challenge to receive expert training and the
21:09opportunity to qualify for up to $1,000 in practical assessment tools that you can use
21:14to help reduce turnover, workplace drama, and unnecessary labor costs while generating
21:18more discretionary effort.
21:20And you can learn more at workforcealchemy.com.
21:23So now I got that little plug in.
21:25Let's continue.
21:26You've mentioned earlier something about the future is going to belong to people who can
21:31manage AI agents.
21:33Give us a glimpse of what that looks like and why you say that.
21:38So I talked about one type of AI agent that's doing automated client prospect outreach.
21:46So that's a great tool that can be used to the fact.
21:49You might actually step back.
21:50You might step back and describe what an AI agent is because we have people of varying
21:56levels of understanding out there.
21:58So if you wouldn't mind, hey, what is an AI agent and what's involved in managing one
22:01and what kind of things can you do with it?
22:03And you can kind of just give us an overview of what you see coming in the future.
22:07If you had a little crystal ball and say, hey, Mason, here's what's coming.
22:10And what would you tell us?
22:12So when you want to think about an AI tool in general that you're going to use, you should
22:18think of it as a new employee that's just been hired at your company who doesn't have
22:24any of the context for what you do, but is very quick, very fast, a quick learner.
22:31But again, you need to give it all the context and all the instructions for to do tasks effective.
22:36Now, one way you can do work with the employee is tell her or him exactly what to do each
22:42time.
22:43So that's called prompting, prompt engineering.
22:46So prompting is just telling the employee, do this task, do this task, do this task, do
22:51this task.
22:52Prompt engineering is creating the kind of request that's going to effectively get you the outcome
23:00that you want.
23:01So you give the employee specific guardrails, instructions, everything that you want to
23:07be accomplished in a task.
23:08Like, let's say you want the report done or you want to triage your email or prioritize
23:14the tasks in your day.
23:15You want to give the employee, of course, very specific instructions about what does it
23:19mean to triage your email?
23:20What's important?
23:21And so on.
23:22So just talking to the employee to give it specific tasks, that's prompting.
23:27Prompt engineering and prompt engineering is creating a prompt that's going to be really
23:31high quality.
23:32So those, that's prompting and prompt engineering.
23:34Now, the next stage.
23:36I'm going to interrupt you one second there because something's jumping out at me and I
23:38don't want to lose it.
23:39And I hope I don't throw you off your rhythm.
23:42But I found that taking the time to improve my competence in prompt engineering has made me
23:49a better leader and communicator with real humans because it is teaching me not to assume that
23:56someone understood what I wanted or that someone can read my mind.
24:00I have become much more clear in my communication with vendors and team members about,
24:06hey, this is exactly the outcome that I want.
24:09Here's what I would like you to do.
24:12Here's how I want you to do it.
24:15And here's why it's important.
24:16Now, AI is probably not going to care about why it's important, but humans do.
24:20But I have become much more conscientious about the quality and the detail that I give
24:27instructionally to people that I work with because of this.
24:30So trying to work with AI has made me better at working with humans.
24:34But anyway, go ahead.
24:35Please continue.
24:37That's a good example.
24:38Yes.
24:39So that's a good way of thinking about it.
24:41That's exactly the right mindset to give it very clear, thorough instructions.
24:44So that's prompting.
24:45So is talking to a prompt engineering is creating a clear, specific request that gets you what
24:51you want every time.
24:53The next step is agents.
24:55That means taking an engineered prompt, one that gets you the requests that you want every
25:00time, and having it take actions in a scheduled way in the real world.
25:06So, for example, we talked about the example before.
25:10So, first, how do they work with the company to build that prompt that would be reaching
25:16out to people?
25:17Or how did the company build that prompt?
25:18So, the company build the prompt by the engineered prompt by first figuring out, okay, what needs
25:26to happen?
25:27There needs to be a number of activities.
25:28We need to identify prospects, these triggers.
25:32So it needs to go online and research announcements about divorces and about business acquisitions
25:37in the local area.
25:37Then it needs to take that information and customize a pitch email to them using the company branding
25:45guidelines and structure and template of how to create a good pitch email to them.
25:51And then it needs to draft that email.
25:54So not simply put it in the chat, but put it in your email ready to send.
26:01And then people would actually send.
26:04And the next last step should have been people editing it to be in their voice, which again,
26:10we talked about how that was an issue.
26:12So that is an agent.
26:15It takes actions in the real world by researching.
26:17It then drafts the email internally, and then it puts it in your email draft file.
26:25So in your email draft folder.
26:27And so that's one type of agent.
26:29And there are many, many types of agents that there could be.
26:32So for example, one of the agents that I just did a training this morning for a company,
26:39and we created an external meeting prep agent.
26:41So what would happen is for meetings with prospects and any other external meetings with external stakeholders,
26:49with vendors, so any external meetings, it would gather information from the company's CRM,
26:56from the company's, from the email exchanges, from the calendar, from the company's SharePoint,
27:03from the company's OneDrive, and other company files and folders.
27:09And then it would draft a meeting pack.
27:12So a preparatory meeting that would email to the person who is going to have the meeting an hour before
27:18the meeting.
27:20And so the person now doesn't have to do any of that prep work,
27:23which used to take a bunch of time preparing for meetings with prospects, with clients, with vendors.
27:29So now that's done.
27:30So that's another type of meeting of agent.
27:32So again, that's something that looks at your calendar and is able to, based on the calendar,
27:37create all of that information and mail it to you at the right time.
27:40So that's what agents are.
27:42They're able to take actions in the real world, drawing in different data sources,
27:46following consistent, specific instructions.
27:50And so in the future, everyone will be managing a suite of such agents
27:56that address a whole variety of workflows that they're currently doing.
28:00So instead of doing all these tasks themselves,
28:02they will be managing these agents, making sure that these agents are functioning correctly
28:08and that they're providing them with the information they need
28:11so that the salesperson doesn't have to do any outbound information.
28:15They're only taking inbound messages.
28:17So that's like, well, just going in a deep, deep run sales.
28:21So they're completely prepared for any meetings.
28:23So that with clients who they currently have relationships with,
28:27they have automatic follow-ups that are going to be customized
28:30and ready for them to send.
28:32So they're going to be draft emails.
28:34So lots and lots of things that make them super efficient
28:37and focused on the things that really matter.
28:39With salespeople, it's going to be making the sale
28:43and then retaining the client over time.
28:46And so that is what agents will be,
28:49what the agents already are,
28:51and they'll get better and better over time.
28:54You have me imagining the future and my instincts.
28:58And we'll see when we look back years from now
29:01and reflect on this conversation, whether I was right or not.
29:04But I can see the need for developing the skill
29:09to manage the agents, just like you've described.
29:11I don't think the ability to lead or manage humans
29:15will ever become obsolete.
29:17I can see a short period of time
29:20where there is perhaps a trend
29:23where people are going to overdo it on,
29:24hey, I need to learn how to manage all these prompts
29:26and all this AI.
29:27And they may neglect the soft human side of it.
29:31And I think that will come back to bite people in the end
29:35when the pendulum swings back towards the balance
29:39between leading people and managing agents.
29:43And I'm curious to see how the future unfolds.
29:46But my gut tells me that,
29:47yes, people now that have aspirations
29:51of management and leadership and business
29:53are going to be well served
29:55to learn how to manage agents.
29:59And I think the ones that go really far,
30:01the best of the best,
30:03are going to continue to focus on developing
30:05their interpersonal skills
30:06and their leadership skills as it relates to people
30:08because that's truly the best of both worlds
30:10for people that I think have that vision
30:13of being the best of the best
30:14in whatever industry they're in.
30:16We'll see if I'm right,
30:17but I'm willing to go on the record right now
30:19and say I think that's what's going to happen.
30:21We'll see.
30:22Yeah.
30:22So I talk about that in the later chapters in my book.
30:26I don't think it's going to be that flat,
30:31kind of managing people.
30:32It's not going to be that.
30:33The key skill of the future
30:35will be managing people who manage agents.
30:37I see that.
30:38So it's not going to be like flatness.
30:42It's going to be managing people who manage agents.
30:44That's the skill of managers in the future.
30:46And so I talk about that in the last chapters of my book.
30:48Like I said,
30:49there's the College of AI Adoption at Work
30:50from Resistance to Results.
30:52And so the skill would involve figuring out,
30:55okay, you're leading, let's say,
30:57a team of eight people.
30:59You have each person is managing a suite of agents.
31:02What's happening at the handoffs?
31:03Let's say you have a salesperson who made the sale.
31:06What's happening at the handoff
31:08from the salesperson to the account manager?
31:11How is that going?
31:12How is her agents figuring,
31:16you know, Mary's agents who is making the sale
31:19to Bob's agents who are managing their account, right?
31:22How's that going?
31:23And all of these handoff,
31:25team collaboration issues.
31:27How do you effectively manage people,
31:29a series of people,
31:30who are managing their own agents
31:32and how do you get them to work together effectively
31:34while also managing your own agents?
31:36So, for example,
31:38with managers,
31:40they need to learn
31:41how to have different types of agents.
31:44So, for example,
31:44they need to learn
31:45how to manage performance management agents.
31:48Ones that go across a year,
31:51let's say once a year,
31:53annual performance evaluation,
31:54they take the data
31:56from across a year
31:57and put them
31:58into
31:59an AI database,
32:01let's call it a project,
32:02with all the information
32:04about someone's deliverables
32:06over the course of a year
32:07with instructions
32:09on how to evaluate
32:10those deliverables
32:11and create an ongoing
32:12performance evaluation
32:14so that you don't have to evaluate,
32:16you can evaluate them
32:17with a thorough performance
32:18once a year,
32:19but ideally,
32:20you'll be evaluating,
32:21you'll be sending them
32:22a weekly performance evaluation
32:25automatically
32:25after, of course,
32:27you edit it and tweak it
32:28so it creates
32:29a weekly performance evaluation
32:31email draft for you
32:33which you can edit and tweak
32:34before sending it onward
32:36to this person
32:37about how you've been doing
32:38the last week
32:38or whatever else,
32:39whatever cadence you want to set.
32:40And so,
32:41that is a different type of agent
32:44that's specifically
32:45for managers.
32:46So, managers need to learn
32:47how to manage their own agents
32:49that are going to be specific
32:50for managers
32:51and they need to learn
32:52how to manage effectively
32:53people who are managing agents
32:55that are ensuring
32:56good team collaboration
32:58and handoffs to each other
32:59and addressing that last issue
33:01of those,
33:02the secret graphic profile,
33:03I talk about that
33:04in Chapter 6,
33:05of that cautious,
33:06reluctant adopters
33:08who are,
33:09who feel shame and stigma
33:10around using AI
33:11and talking about it.
33:13Let's talk about those people
33:15for a second.
33:15How can leaders
33:16encourage employees
33:17to maybe experiment with AI
33:19while still protecting
33:20sensitive information,
33:22quality standards,
33:22and customer trust?
33:23I mean,
33:24I almost like to think of
33:25like a kid in a sandbox.
33:27It's like,
33:27play with it,
33:28see what happens,
33:29and what's your thoughts
33:30on that?
33:32So, to ensure
33:33that they do that
33:34and they talk about it
33:35and they're not using it
33:36in a shadowy fashion
33:38and that the team
33:38gets benefits,
33:39what they need to do
33:40is, first of all,
33:41model it themselves.
33:42So, that's critical.
33:44Leaders need to talk
33:45about how they're using it,
33:46how they're using it
33:47for performance evaluation,
33:49how they're using it
33:49to draft their emails,
33:50to draft the reports,
33:52to do financial models,
33:53to look at profit
33:53and loss statements,
33:55analyze those
33:55and create those.
33:56They need to be talking
33:57about how they're using it
33:59and getting,
34:00addressing the social stigma
34:02of using it
34:03to make sure
34:04that that is addressed.
34:05That's the first
34:06fundamental.
34:07Second,
34:08they need to provide
34:09a baseline of training
34:10to everyone
34:11and, again,
34:12on the tasks
34:13that people dislike to do
34:14but they don't want to do
34:16so that baseline
34:17of training
34:18is going to be there
34:19and that people
34:20can talk about,
34:21hey,
34:21I'm using AI
34:22to do these things
34:23so that there's
34:24public usage.
34:25Then,
34:26they need to provide
34:27an asynchronous venue
34:28for people
34:29to actually talk
34:30about their AI usage.
34:32So,
34:32have a Teams channel,
34:33a Slack channel
34:34where people share
34:35about their AI accomplishments
34:37and what they're doing
34:38effectively
34:39and the leader
34:39needs to positively
34:40reinforce
34:41and reward them
34:42with words
34:43of encouragement
34:44and I encourage
34:46companies to set up
34:47like small financial
34:48rewards
34:49for especially effective
34:50time-saving techniques
34:51that save companies
34:53time and money
34:54and improve
34:54their resources.
34:55So,
34:56thinking about those metrics
34:57that are going
34:58to accomplish
34:59your outcomes.
35:00So,
35:00ideally,
35:01you would do that
35:01but at least
35:02you can praise them
35:02for doing it.
35:03So,
35:03those are the three elements
35:05that are really helpful
35:06for addressing
35:07the reluctant adopters,
35:09getting them,
35:10addressing the social stigma,
35:12addressing that shame
35:13and making sure
35:14that you have modeling,
35:15training
35:15and post-training,
35:17sharing
35:18and rewarding,
35:19raising
35:20the usage
35:21of AI tools.
35:23So,
35:23if there was
35:24just one piece
35:25of advice
35:26that you could give
35:27to business owners
35:28or CEOs
35:29or executives
35:29that are listening
35:30or watching today,
35:31what would be
35:32the most important
35:33piece of advice
35:33you would offer them
35:34and why?
35:36It's not the technology,
35:37it's the psychology.
35:39The technology
35:40is actually quite,
35:40quite easy to use
35:41once you learn
35:43how to use it.
35:43You need to address
35:44the psychology,
35:45you need to address
35:45the three big
35:47emotional blockers
35:48that we talked about.
35:49Fear and anxiety,
35:50identity threat
35:51and shame.
35:52Once you address
35:53the psychology,
35:54once you flip the switch
35:56for people's mindsets,
35:57they can figure out
35:58how to use AI tools
35:59with appropriate training,
36:00of course,
36:01but even without the training,
36:02it's very easy
36:03to learn how to use
36:04the AI tools.
36:05So,
36:05that would be
36:05my one piece of advice.
36:07Remember,
36:07it's not the technology,
36:08it's the psychology.
36:09First of all,
36:10thank you.
36:10If people want to know
36:11more about you,
36:12your books,
36:13your work,
36:14the consulting
36:14that you do,
36:15what are the best ways
36:16for them to connect?
36:18Well,
36:18the first thing
36:19is to reach out.
36:21The best way to connect
36:22is on LinkedIn.
36:23I'm very present there
36:24and you should tell me
36:25you heard me on this podcast
36:27because I don't just accept
36:29LinkedIn connections
36:30from anyone.
36:30There's way too much spam.
36:32So,
36:32they'll look up my name,
36:33Dr. Gleb Ciporski,
36:35on LinkedIn.
36:36Connects with me
36:37very easily available.
36:38That's the first thing.
36:40But if you want to,
36:41and that's connecting with me,
36:42if you want to learn
36:42more about my work,
36:43so if you want to get
36:45a free introduction
36:46and chapter from my book,
36:48go to
36:50disasteravoidanceexperts.com
36:51forward slash
36:52AI book.
36:53Again,
36:55disasteravoidanceexperts.com
36:56forward slash
36:57AI book.
36:58You'll find the introduction
36:59and a free chapter
37:00for the book.
37:01And if you bought the book
37:02and you put your order number in
37:04from Barnes & Noble,
37:06from Amazon,
37:07or a receipt from
37:08a physical bookstore,
37:09anything like that,
37:10put it in
37:10and you'll get
37:11a free assessment
37:12associated with
37:13the book
37:14on adopting
37:15AI effectively
37:16in your company
37:16and a manual
37:18on the seven mistakes
37:19that leaders tend to make
37:20with AI adoption.
37:21So it's all going to be
37:22at
37:24disasteravoidanceexperts.com
37:25forward slash
37:26AI book.
37:27Dr. Gleb Ciporski,
37:28I really appreciate
37:29your time
37:29and your talent
37:31and your expertise.
37:31Thank you for sharing
37:32that with us.
37:33And before we wrap up,
37:34I just want to leave
37:35people listening
37:36with one practical
37:37next step.
37:38If you want
37:38less turnover,
37:40less workplace drama
37:40and more discretionary effort
37:42from your team,
37:42apply for the
37:43Workforce Alchemy Challenge
37:45at workforcealchemy.com.
37:47You'll receive
37:47expert training
37:48and have the opportunity
37:49to qualify for up
37:50to $1,000 in tools
37:51that you can actually use.
37:53Thank you so much
37:54for joining us
37:55and it's been a pleasure
37:56having you,
37:57Dr. Ciporski.
37:58Thank you again
37:59for inviting me,
38:00Mason.
38:00It's been great.
Comments