Thriving in Engineering

Carol Palombini is a Leadership & Culture Consultant specializing in the human side of AI adoption. A psychologist, former Head of HR, and executive coach with extensive experience working with technology leaders, she helps organizations assess whether their leaders, people, and culture are ready for AI-driven change.

Her work brings leadership readiness, psychological safety, workforce attitudes, and burnout risk into AI adoption strategy, helping organizations create the conditions for AI to move from experimentation to sustainable adoption.

This is the Thriving in Engineering Podcast, where we talk to engineers and technical leaders about what AI is actually changing in day-to-day work, past the headlines and the hype.

In this conversation, we discuss:
  1. Why AI adoption resistance is an influence problem, not a technical one
  2. The two types of leaders AI disruption is creating
  3. Why a three-year AI vision is impossible for AI adoption
  4. Why metacognition is becoming engineers' most valuable skill
  5. Why measuring AI adoption by tokens burned backfires
Carol's details:
Website: https://www.carolpalombini.com/
LinkedIn: https://www.linkedin.com/in/carolpalombini/
Substack: https://carolpalombini.substack.com/

What is Thriving in Engineering?

Software engineering isn't just strategy and sprint plans. It's ambiguity, pressure, trade-offs, and people.

Thriving In Engineering helps you navigate all of it.

In the newsletter, Alex Ponomarev distills 25 years of engineering leadership into practical frameworks to help you lead with confidence, stay calm under pressure, and build thriving teams.

The podcast goes deeper, talking to engineers, leaders, and builders across the industry about what AI transformation actually looks like in practice, past the headlines and hype. One guest at a time.

Hi everyone.

my name is Alex Ponomarev and welcome to Thriving in Engineering Podcast.

In each episode, I talk to people from across the software industry about a my guest today is Carol Palombini.

Carol, welcome to the podcast and thank you for joining.

Please tell our listeners and viewers a bit about yourself.

Thank you for having me, Alex.

So my name is Carol Palombini.

I'm a psychologist and tech leadership coach who works with leaders that already delivers a lot of value, but still struggle to show their value.

So I work with influence and visibility for tech leaders.

Thank you.

This is amazing.

We've we obviously live both in Lisbon.

We met at the Engineering Leaders Conference in Prague in April.

You were giving a keynote there, and you're also doing roundtables for engineering leaders.

Can you tell a bit about what you're teaching people, coaching people about?

Of course.

Well, my talk during the ELC in Prague was mainly about one of the patterns that I see like hindering influence among tech leaders and especially executive ones, but also

middle management.

I think that there's a mindset issue that we have to address.

And I've been talking about this mindset issue and the patterns that I see across the the tech leaders' mindset.

that are hindering their growth or hindering their visibility.

So in the at the ELC I was talking about the lack of organizational awareness and how that hinders their ability to read the room.

How it's difficult for usually for engineers and for leaders to have analytical people in the end to have a strategic view and read people's intentions and understand how decisions

are made.

So they really

count too much on the right and wrong perspective.

They're really trying to push some argument to showing their right.

And in the end, that can even take them further away from the decision they want people to make from them.

So at the ELC was that, but mainly in the round tables we discuss any type of influence and visibility issues that leaders may have.

So they can choose a topic.

Usually the regular round tables run like this.

They choose

The mini groups, each one choose that's a hot seat.

So you choose someone's problem to tackle and help with.

But most of the problems go around, I want to have more say in the strategic table.

The strategies come to me already done, and I have to I'm at an execute.

I'm seen as an executor instead of being seen as a strategic leader.

And that has everything to do with the lack of ability of reading the room and understanding how to influence people.

So it's mainly about that.

This is amazing.

I think everyone well, I don't know about everyone, a lot of people can relate to this.

I definitely can, because I'm a former engineer and transitioning to management was a huge challenge precisely because of that.

Like as as an engineer, you're always focused on on the solution, on the right solution.

Whereas strategy is not necessarily about the right solution from the technical standpoint.

There there's this bigger picture.

There's corporate politics and like long term planning, emotions as well, that we as engineers don't pay enough attention to and we need to learn.

And it sounds like that's exactly what you're helping with.

I saw the topic of the latest round table on LinkedIn.

That one was about the AI adoption, and that's

That actually aligns a lot with with the topic of of the podcast.

Can you tell a bit more about how that round table went?

I'm personally super interested.

It was surprising.

I have to tell you that I had some outcomes that I wasn't expecting, but was super rich anyway.

this was a special edition because usually, as I said, they pick the topic they want to address.

But this time I think I wanted uh to create a different dynamic so people could address the same issue.

But in the end, it turned out they weren't addressing the same topic.

So the topic was resistance to adopt AI by the team.

So why the reasons engineers, some engineers are resistant to adopt AI and how you can help them overcome this resistance as a strategic leader, what is your role in the end?

So what I wanted to explore there was we they had two moments of group work.

The first one was okay, so if this is an influence issue, because AI adopts adoption is an influence issue.

It's not a technical issue.

You have to convince your team, your stakeholders, the board, the leaders, everyone of your strategy.

The best way of doing this, of adopting this.

So the first mo the first part is if you want to influence, you have to empathize.

So let's empathize with the these stakeholders we are addressing today, which are the team members, the engineers.

And so what's in it for them?

Why would they

adopt AI if they fear so much being replaced by it?

Okay, so then the groups went, they divided into these mini groups, but when I joined some of the groups and then we got back to discuss the results.

So what I noticed was some of them really got to the discussion of the benefits, but some of them couldn't even get to the discussion because they stepped back into

This is not the right thing to do, a to adopt AI.

Should I adopt AI?

Or complaining about the the other C levels or the board who is making a lot of pressure.

So what I felt at that moment is that some leaders are really afraid of these.

So there are two types of leaders.

There are two types of leaders dealing in dealing with this.

I think there are the ones that say, okay, so this is life.

It doesn't matter if I like it or not.

This is what we have for today.

This is my next challenge.

I'm going to tackle it the best I can.

And they are still struggling with the how.

But I feel that I at least my impression was, and I would like to to hear from you as well, is that most of them are skeptical.

They are judgmental of the reasons for for adopting that, and they are afraid of themselves not knowing how to deal with that, or maybe afraid of being the bad guy in this

situation.

In the end, the person that's going to make people redundant, that's going to end all of these the people's dreams.

Shatter lives and shatter so they are so concerned about being in this role that they wouldn't like to be that they don't allow themselves to actually get into the discussion,

get to the point of what they can actually do, what's in their power to do.

So that was like the first thing.

Did do you feel like the the same have you been feeling this fear?

Also until

thank you for asking me.

Well definitely I I guess in my team and and I'm leading a team of twenty five, maybe thirty people, for us it's a bit different because we're actually past that

stage.

We were pushing hard for AI adoption maybe a couple of years ago, just because the field we're we're working with, we're we're working with very

forward thinking companies like venture capital firms.

And so we're somewhat ahead of the curve.

But even for us, it it definitely was a challenge.

People like the engineers, they were scared, very emotional.

They divided into camps.

Some were overly optimistic, the others were super skeptical and didn't want to try things out.

And it's interesting for me that

a lot of companies are still at a stage where they're debating whether they should adopt AI or not, where you know, for a lot of people in the startup world, a lot

of people who are, you know, so ahead of the curve, for them it's this is no longer the question.

For them, it's like how do we measure uh the progress?

What do we do next?

What what exactly do we do?

So, you know, adopting or not is no longer the question.

I think it's such a rich topic.

I wonder like what were the conclusions?

Were you able to help them to kind of move past this?

So the second part, so this was the first part, and there was the discussion and in the end they come up with some of them came up with the list of things they thought it was

beneficial because for engineers, because the the whole point was if you want to influence you have to speak directly at people's desires and dominant mind reasons,

what motivates them.

So speaking to their hearts and and removing fear.

It's impossible to speak to the hearts of people if they're blocked by fear.

Fear is like a cloud that's it's in front of your eyes all the time.

If you don't remove it, you won't be able to see anything, not even the benefits.

So it's important to do that.

Some of them can so the second part, they had to discuss their role.

So what is my role as a strategic leader in the in in an I adoption?

So what is the the my vision for this?

That was completely crazy because they there was only one conclusion that they got.

It's that's impossible to create even a three-year vision for how AI adoption will look like, for what these roles, like the engineering role, would look like.

So for them, this is creating a lot of anxiety.

Tech mines, they thrive on certainty.

They thrive on being right.

And this is creating like a mess in their mind.

So if I cannot deliver this concrete vision of future, they don't know what to do.

So we debated a lot, and I was showing them that not all vision should be a technical future.

You can have a vision that more that's more based on empathy and building together.

So if we are living disruptive times, you can compare that with I don't know, your sailing.

On a sailboat, like you have no motor, you depend on the wind, the ocean is completely messy, and you have to navigate that craziness.

All you can do, you don't know where you're going.

So the all you can do is like let's row in the same direction.

Let's take let's be together, let's build this together.

So it's more the vision that you have to have as a strategic leader now that will

influence your team, bring them together.

It's what if we are the ones that get this right?

We don't have the answer of how is the best way to adopt AI?

And I imagine that this is very particular to each company.

There's not gonna be a right and wrong way, there's not gonna be uh one single way of doing this.

But what if your vision as a strategic leader were only to was only to say to people,

We're going to get through this together.

Because that was the only thing that they could all agree on.

That this should be co-created with the teams.

This new vision of engineering.

Because one of the things that's bringing a lot of anxiety is there's also this this polar thinking regarding what coding will look like.

So some people believe, and I saw this from the round tables and the other conversations and had, that AI will replace coding, all coding, and engineers will become

something else.

So that poses a very interesting threat for engineers because engineers nowadays are super connected to being coders.

They are proud of that, of being like human machines and actually being

craftsperson.

So they they were praised, they were rewarded their entire lives for doing this, for being this.

So they are their identity is connected to this craftsmanship.

So and so what do you have to do if you do believe that coding will eventually be done completely by AI is help your engineers go through this and reframe

the way they are, rethink, reimagine their roles as engineers.

So what will this look like if you are not doing the coding anymore?

But there are those leaders who think the opposite.

They think that AI will create crappy code and that we will always need engineers to coding to either to fix that or to understand what AI is doing.

So those are completely different visions

of this future.

And the pessimistic vision of this future is definitely getting in the way of finding solutions.

Yeah, I can agree with that.

I believe that while those are two polar visions, the reality will be somewhere in the middle.

And you know, most code will be written by AI.

We can see that today, but it doesn't necessarily mean that anyone is getting replaced.

But what's interesting in what you just told is that a lot of engineering leaders

have trouble with uncertainty.

And I again I can relate to that because when you transition from an engineer to a leader, one thing that hits you right in the face is total lack of predictability of

anything.

Because when you're an IC, somebody tells you what to do.

Everything is predictable.

You do one, two, three, then somebody gives you another task and you can continue.

When you become a leader, you're

Just like you said, you're in the boat in the middle of the ocean.

You know roughly the direction and you have to figure out.

You don't have, you know, no motor or anything.

So you have to improvise.

And I think AI just amplifies that because there's just much more uncertainty now than before, because good luck playing for three years, but it's hard to plan for three

months because things are changing week by week.

And I think that that demands a different framing of planning.

So it instead of planning for tasks, which is also another a trait of tech minds of analytical people, is to be task oriented.

So they're super concerned with the tasks.

And this is not the time to be concerned.

This is a time to reinvent and reimagine this and to rethink this.

But you can plan on other things.

You can plan on the the growth of the team.

Of the level of debates that you want from this team.

And I feel that's missing if you know that AI is going to do a lot of the hard work, what is the work that's human that human that really needs humans and that you should be

planning on evolving these people, on developing these people to get to the level that you need?

Like, for instance, with metacognition.

Metacognition will be, or it is already, the most valuable skill.

An engineer can have in the AI disruptive era.

It's the ability to actually think about your thinking process.

So if you are able to think about your thinking process and translate that and communicate that clearly, you will of course get the best results from AI, train the rest of the team.

So it the planning now should be more in terms of readiness for uncertainty.

If my boat is too big, how can I make it

Smaller so I can shift faster.

So it's more about planning for uncertainty, not planning, expecting this plateau that will eventually come.

And we don't know if it's going to come, but it it because if it's not AI, it's a war, it's an pandemic, it's so uncertainty seems to be here to stay.

So if you're waiting for

something to solve, I don't know who would solve and what you you expect to be solved as a way of using AI, or once this is established, something else will come after this.

So actually building resilience in teams to deal with that is a challenge for tech leaders because tech leaders have this tech mind and the teams also have.

So everyone hates this.

Everyone hates uncertainty.

How are you going to deal with that from now on?

Yeah, I think we we should start loving uncertainty and thriving in it.

That's the only way.

I think I saw something like an event you did with CTO Craft, maybe another round table that was titled "What got me here won't get me there."

And I think it's like exactly this topic.

If you were an expert before, and you were really good at going deep, now you got promoted, you're a leader now.

Good welcome to the new life.

you have to learn new skills, basically.

And people skills, not just any skills.

The skills you didn't want to learn your entire life, and then suddenly they ask you not to be too technical and not and just to pay attention to people and not that I don't

don't mean just the teams because I actually think tech leaders are very careful with their teams.

They pay a lot of attention.

It's their team.

It's like this, my team.

It's not like okay, now I have to have a more less focused view, more broader view.

And I can I have to pick on people's expectations.

I have to map these people, these the these people expectations, what they expect from me, and not only know what they expect from me, but say it back at them, confirm at them,

agree with them, and manage expectations.

So the first expectation, every leader that's stepping into this more strategic goal that understood the technical prowess would not take them to the next level.

It's understanding your own expectations.

Who do you want to be?

What is your leadership mission in the end?

So what are you proposing to implement?

And that's another thing that I think is it's a gap in tech leadership.

Too reactive.

Tech leaders are actually waiting for the CEOs, CFOs, CXOs.

to tell them what they expect from them and to tell them what to do.

But when you are at this level, you have to propose.

You are the highest authority in technic in technical issues.

So you need to be the one who propose.

You need to ask them what they expect, what they want as results and propose to them.

And I think that has everything to do with AI adoption.

most leaders are complaining about things that are demanded from them, but most of them had never presented their proposal, their view of how this should be done.

So because when you do that, you also have to say, Okay, so this is what expect happen.

Let me be, let me do my tasks, and I'll bring you the results that you expect.

That's a lot of responsibility as well.

You cannot blame anyone else.

Yeah, a hundred percent.

And I think there is an important piece to it.

your CXOs and stakeholders you work with, they should let you drive this.

They should believe in you and they should give you like hand you this responsibility.

As a leader, you have to have this influence, you have to have this authority.

Maybe as you are starting out, you don't have it initially have to build it up.

And you have a lot of experience working with people, managing people.

You were the chief people officer in a large Brazilian company, right?

Yes, I was chief people often three companies in Brazil with large projects, medium companies, but with very large projects.

For once, one of the companies I worked with was doing the Confederations Cup in Brazil.

So it was responsible for all food and beverages in all the venues.

So it was insane.

But mainly I was head of HR in a tech company as well, was the last company I had this role before I became fractional and did all this work for as a consultant.

But

I felt this is what I feel about tech.

That's why I chose to work in with tech people, because I think that in the end they are very good leaders.

They they have good intentions, they're not competitive tech leaders.

They they really want to co-create with the team.

They have the best intentions.

And the problem is just going through intention to behavior.

Because they assume too much.

They assume the team knows they value them.

They assume that the the CEO trusts them or will

come to them if they have a technical issue.

It's too much, too many assumptions.

And you have to jump from actually communicating to people your expectations.

And as you mentioned, the CEO should give the authority to the CTO when they join.

This is definitely the case.

If you don't have the buy-in from your CEO from the start, there's something wrong.

He just hired you.

Why wouldn't he buy in?

And what you have to do is actually make that buy-in very transparent to the entire organization.

You have to ask them the way you want to be presented.

You have to ask them if you want if they allow you to use their name in certain projects or tell them, Okay, I have the buy in or Alex is with me on this.

Talk to Alex and we decided that we are going to go with on with this direction.

yeah yeah

That and that nothing works and and and and I see that from experience w when the CEO is not aligned with one of the other C levels, nothing goes well.

Yeah, yeah, hundred percent.

Here's a question for you related to this.

a lot of companies are thinking about AI adoption, AI transformation, tech teams are adopting new tools, coding agents.

but it seems like there's this ongoing debate of how do you actually measure the success of this transition transformation, especially now that a lot of

huge tech companies are reporting negative ROI in after spending years on this transformation and a ton of money.

So I'm curious what's your stance on this and have you heard opinions of you know of the leaders you work with?

How do they approach this?

Well, I haven't actually gone into this subject and but this is a very good one.

I will take me I will take it to the next round table.

But I think that everything that has to do with assessment of of success, any type of assessment of success, you have to understand the first thing is that the way you will

measure will change people's behavior.

And that's hardly how people think about performance reviews and about success.

So if you think just in terms of financial

and money and the costs this will also make people feel less valuable and you will have a a bigger churn in the end so you have to make an equation that will

lead people to adopt the tool that you want thinking about how will I how can I motivate people through these uh these KPIs

What is the backlash of doing this?

Because all of them have.

It's almost impossible to create a KPI that will measure success success without a backlash.

So you have to look that deeply into that.

What is the outcome of this, the good and bad outcome of this?

What's going to be motivating and demotivating to whom?

And how do I minimize those impacts, those bad impacts?

Maybe it's creating a second KPI that balances this one.

So if you're thinking about a you the first thing you have to do is testing, I think.

If you if you're going to start using something, you have to think this, and I think that that's the one of the mistakes companies are making.

They're not taking it as a test.

They're not they're not doing any pilots.

They're going all in.

So so when you go all in, there's one only one way of winning, right?

If you win it all.

So if you actually get that hypothesis that you were using.

Like all coding will be done by AI.

I don't I can replace people with AI and then you will have to go back.

There's just either or.

But if you take this as a test, you can have many types of KPIs and measurements that will not demotivate people to try again.

Mm-hmm.

Yeah.

Can't agree more.

To me personally, this is one of the most interesting topics in leadership because just in, like on AI adoption, there are two polar approaches.

One approach is, well, we will set up a very specific KPIs, and that could lead to the Cobra effect where you you measure the wrong thing, it changes the behavior.

And for example, measuring tokens,

in many companies they measure AI adoption by how many tokens the teams burn.

So what what's happening is people start to burn a lot of tokens.

Companies spend a lot of money on those, but there's no real progress in terms of productivity growth.

The other opposite is measuring the feeling.

Well we will measure if we feel more productive.

And then this is very subjective.

The engineer may feel more productive and then the CEO doesn't feel it.

And it's also wrong because it's subjective, it changes from day to day.

And finding this there's no rule book to find the right balance.

Just like with most of things related to people, there's no specific grades or something.

You have to find your own balance for your own team, for your own situation.

And so company and each company will find their solution as long as they work with teams.

The best people to tell you how to measure, it's them.

They will know.

They will know.

We we all know when you are starting over, you always complain about things that should be measured and it's not, things that you think that are more valuable about your

work and the company never measures.

I remember that as a recruiter when I was starting my career as a recruiter, a tech recruiter.

I was like, oh my God, why don't they just care how how long we take to to close this opportunity, but not the quality of the candidates we are delivering, for instance.

So the options they have.

So if you talk to them, you can workshop this with them and try out different things.

You don't have to to get it right the first time.

You can try out, see if it works.

But in my opinion, and this is really my opinion, I think that

In terms of AI, you can measure before and after.

What was working well before?

Is it working better now?

Like or are if you're talking about if your product is related to customer relationships, are customers more satisfied with with us?

Or if you're talking about SLAs?

So is it reduced?

Are are we faster?

Is the quality better?

So you can do before and after, don't you think?

Hundred percent.

I like the the measurement of happiness in CSM and customer success management.

There is a very old anecdote where a company hires a guy who's who he claimed he's very fast at at math.

Yeah.

So they hire him to to work as a mathematician.

And then sometime later that they come back to him and say, Hey, we got the numbers from you, but they're all wrong.

Well yeah, of course they're wrong.

I never claimed that I'm do I'm good at math.

I'm just fast.

So a lot of adoption is about doing things fast, but not not necessarily doing them right.

I think the rush in some cases, those cases that we saw like Duolingo, they rushed into firing a lot of people.

Now they have to rehire a lot of people because things didn't go as expected.

I don't know, but I feel like some CEOs must hate human beings.

Like they just want to get rid of people.

And if they could, they would be the only humans in the company that will be entirely run by people by machines.

So

What's that all about?

It's just like the cost thing, just they just believe that they're they're going to do something more costly efficient, or don't they they really don't want to handle humans?

What do you think?

I think there's a lot of pressure on the leaders.

There's a lot of pressure on everyone.

We're in this transition stage that really comes once in a lifetime and and not every generation is lucky to have lucky, quote unquote.

to have a transformation like that.

It's almost like the invention of electricity where nobody knew what is this new thing, how it's going to work.

How it's going to change things, is it safe?

Is it gonna kill us all?

and so on.

So there's there's tremendous amount of pressure.

And based on my experience, the high like the more people you have that you're responsible for and you report to.

And if you're a CEO of a large organization, you have a lot of people you're responsible for.

You're you have a lot of people who've put their

money and future of their families into you.

And you have to like almost every move you do hurts someone.

So I think for us it's easy to from our perspective to to judge these CEOs but in a way they're they're trying to do their best.

It's just it's a it's a tough spot to be in.

And and I think a lot of higher order leaders they grow thick skin and they just do whatever they have to do.

But what concerns me more is this following like almost like a cargo cult.

If that company did that, we should do the same.

And that I don't really understand.

It doesn't make sense to me because, for example, if you're a if you're in a restaurant business and you want to run or open a restaurant, you wouldn't be copying

McDonald's because McDonald's is a huge corporation.

There's a brand, they have a ton of money.

If you're a small restaurant owner, you can be doing things like McDonald's.

So you have to find your own way.

Just copying them doesn't make sense.

But what's happening now, because of a lot of claims that we can replace engineers, we can replace this, radiologists, this and that, a lot of leaders at smaller

companies are thinking, well, my senior engineers are skeptical.

We should fire them all.

Claude Code or Cursor will do their job.

And it's just it doesn't make sense.

It's it's much more complicated and interesting than that.

But everybody's super scared, and where people are emotional and under pressure, all all kind of weird things are happening.

I think that that's the point is people are reactive because they they fear like this pressure as you mentioned to be like the first ones or to beat the other ones because this

could be something disrupted, could change everything and they're rushing into their being reactive instead of like actually stepping taking a step back and assessing what's

best for the company of course some might have done that and just made a mistake.

But

Some of them like the smaller companies, they don't know where to go and this they are feeling lost.

So are they are using these bigger companies as a North Star.

So if they did this, it's probably the good thing to be done.

And so let's do that.

Although they might not even need AI in their in their businesses.

Yeah, I personally think everybody needs AI in their businesses.

But I'm biased.

I'm AI built as they say now.

But yeah, it how how you actually apply it, where you make the changes in the organization, that's that's very specific to each company.

Carol, if there would be one thing

engineering leaders would take away from our conversation today,

what would that be?

What's your number one advice in this very interesting time we're living in?

Think is have discussions about this.

Don't try to solve it on your own.

Like if you are not ready to take this to your team and co-create them solutions, you have to talk to peers.

Share your fears, share your concerns with other leaders that they are we are all in the same situation here.

Tech leaders are all in the same situation.

It's not about getting things right now, exactly now.

It's about having the discussions now that need to be to be done so you can have a good test.

A good first test to to get this going.

Don't worry about getting this right the first time.

I think that's amazing advice.

In life in general, beyond just engineering leadership, we put too much pressure on ourselves.

Thank you so much, Carol.

It was an amazing conversation.

You gave so much true nuggets of advice.

I would recommend our viewers and listeners to to join your round table to talk to the peers and to understand that

we're not alone in this.

We're we need to support each other.

Thank you.

Thank you very much for having me.

It was a great conversation and I hope to come again.