Trending
Nvidia Unveils DSX POD Amid Greenpeace Protest Nvidia Unveils DSX Pod for Faster AI Data Center Builds Bonus Features – October 18, 2026 – 74% of cyberattacks lead to disruptions in patient care, 56% of caregivers use AI tools to support their responsibilities, plus 30 more stories How AI Is Helping Clinics Do More With Their Existing Staff Physician Collaboration Isn’t Just a Requirement; It’s a Scaling Strategy for NPs Before You Buy an AI Tool for Patient Questions, Count the Questions Abby Care Raises $82M to Scale Paid Family Caregiver Platform TikTok reverses ban on ads for Musk documentary Hands-On, Clinician-Led, and Real: Why DHC2026 Is the Rarest Room in Health IT Another Data Center IPO Fizzles as Investor Appetite Wanes The end of independent infrastructure roadmaps: how AI is forcing data center co-design DeepSeek updates V4.1 Flash without raising API prices How Uber Built a Genie to Answer Questions Correctly Key Healthcare Cybersecurity Lessons From the SHI Summit Outro Announces $7M in Funding to Expand Access to Safe Deprescribing, Starting With Antidepressants

The hardest part of AI transformation isn’t the AI

Google Preferred Source

I come from an extremely tech environment. I live in Silicon Slopes, Utah, and at Neighbor, a storage marketplace backed by a16z and Fifth Wall, our office was open plan. I sat next to the marketing team and right in front of our CPO and engineering, with data and enablement just one row over.

So we were always discussing whatever had just launched in AI, and by the end of the week somebody was already testing it.

So you would imagine the hard part of bringing AI into a company is the technology. After all these years I can tell you the hardest part has always been the human side, the sense of belonging, the respect for each person’s dignity. Pope Leo XIV touched exactly this at the Vatican on October 2, and I’ll come back to it.

At Hallow, I led the Brazilian market from the very beginning until the app was #1 in Brazil on both the App Store and Google Play. Even with AI available for pretty much everything, my decision was always the same: only authentic testimonials, with real people, nothing fabricated.

A prayer app that fakes a person praying is broken at the core, doesn’t matter how good the image looks. I’m especially proud of our partnership with the Christ the Redeemer Sanctuary in Rio, when we projected a message of prayer for Pope Francis on the statue. It went viral, more than 100 million views.

And everything in it was real, the people, the prayers, the moment.

At Neighbor it was the other side of the coin. In hackathons I built solutions together with our engineering team, like crawlers that identified potential hosts by very specific characteristics. Manually, that would take hours and hours, and be full of errors.

And the team got that time back for what humans actually do well: talk to hosts, understand them, build the relationship.

So for me AI has always been a question of knowing which part of the work belongs to the machine and which part belongs to the person.

Now with Fitch & Co, I’m called in to help traditional companies with AI transformation. For the last three years I’ve been working on these projects in wellness, industry and education, and I always start the same way, way before talking about any tool. I start by understanding the culture, interviewing the people, sitting with them.

I come in with more of the right questions than answers.

And curiously, I see a lot of resistance, the same I saw ten years ago with digital transformation in CPG, when I brought influencer marketing into Swift / JBS, or when I was a trainee on Unilever’s first e-commerce team in Brazil, then the second-largest market of a company that today sells over €50 billion a year. The phrases are almost identical, you could copy and paste:

“We’ve always done it this way.” “Our customer is different, this doesn’t work here.” “A machine is never going to understand our business.”

Sometimes there’s a practical reason. The person already saw a rollout go wrong. Or they’re so stretched that “new tool” just sounds like one more thing to learn after hours.

Or they know a problem we, coming from outside, haven’t seen yet. And sometimes the reason is much more personal: if a machine can do what I spent years learning, where does that leave me?

And I understand. I really do. That shield is natural for a human being, a lot of us built our sense of self worth around the things we know how to do.

So how do we end this resistance? What has worked for me is more or less this:

  1. Embrace the human value. Talking to artists and curators, the Pope said that “algorithms lack the spark of the human,” while authentic culture and art preserve it. And that goes way beyond art. It’s in the way a salesperson reads a client, the way a manager notices someone on the team is not ok. So I make it clear from day one that AI is a wonderful tool for managing tasks and saving time. My goal is to give people more room for the creative and relational part of their work. That spark is theirs. I believe a person’s worth goes way beyond what they produce, and in practice that means taking their experience seriously and giving them a real say in how their work is going to change.
  2. Bring it close to the person’s reality. From the very beginning I ask: what do you do today that is repetitive, takes hours of your day or your week, and if you could, you would eliminate or find a way to simplify? People always know the answer. Most of the time they answer before I finish the question.

And when we test, we count the time someone spends reviewing and correcting what the AI did too, otherwise it’s just a pretty number.

  1. Bring the person to be part of the solution. Hands on. We build little by little, together, and then comes the moment I love the most: “look how cool this is, this took hours of repetitive work off your week, and now you can invest that time in new ideas and projects.” When the person built it with you, they see that they got better at what they do and the machine is just helping. The resistance turns into ownership, and often that person becomes the biggest advocate in the company. And for me that only happens when there is trust, when people know you’re on their side and will fight their fights with them.

And here comes a harder conversation. In May 2025, Dario Amodei, CEO of Anthropic, warned that AI could wipe out up to half of entry-level white-collar jobs in the next one to five years. Nobody knows yet if that forecast will prove right.

But if we see this risk coming, what is our role, as leaders, in mitigating it?

One question worries me a lot. If we automate exactly the work where people starting their careers used to learn, how are they going to get experience? We can put newer colleagues side by side with experienced people, invest part of the hours we recover in teaching, and be honest about what we haven’t figured out yet.

I believe that we as leaders have a responsibility here, and AI transformation can’t be only a cost-cutting project with a tech budget, it has to be about people. We can shape the workplaces we influence so there’s room for exchange, learning and growth, with AI as one more part of that ecosystem.

I can’t promise every role will stay the same. Nobody can. But I can take responsibility for listening, for helping people prepare, for thinking about their future when I make a decision.

For me that’s part of being loyal to a team.

The technology we figure out, I promise you. Making people feel they still belong in the room is what takes real leadership.

I’ll leave it as a provocation.


Featured image credit

Tags:

 

Join the conversation

Your email address will not be published. Required fields are marked *