home

Blog

AI culture can't be forced: what we learned by listening to people

Blog, Pitang.

Antônio Valença

AI culture can't be forced: what we learned by listening to people

AI




Everyone uses AI. Few companies use it in a structured way 


We have been working with artificial intelligence at Pitang since 2017. For a long time, this knowledge remained in the hands of a few specialists. In recent years, that changed quickly: AI became democratized and spread, both internally and across virtually every organization. 


Today, almost every company uses AI. The difference lies in how. In many places, it arrives through isolated initiatives, and the knowledge stays trapped where it was born. To generate real results, AI needs to be part of the culture. And culture, as I have learned over these years as a partner and executive director at Pitang, is built with people, day by day, not with announcements. 


That is how our manifesto was born: "The AI tree is a Pitanga tree". In this text, I share what it is, how we address the fear that every change brings, and what our people told us about this transformation. 


What is the "The AI tree is a Pitanga tree" manifesto? 


The "The AI tree is a Pitanga tree" manifesto is the document that guides the adoption of artificial intelligence at Pitang. It treats AI as a capability for everyone in the company, not as a project of a specific department. The motto summarizes the spirit: "Here, AI doesn't tame people. Here, people tame AI." 


We wrote the manifesto because the challenge was never just adopting tools. It was about changing the mindset and making knowledge circulate across all areas, from development to Finance. As stated in the text: 


"The true competitive advantage will not be having a few AI specialists. It will be having an entire company learning at the same time." 


The idea is for it to feel natural, in any area, to ask: "How can I do this better with the support of AI?". And that question is guided by seven principles: 


  1. Real autonomy: each area finds its own path.

  2. Curiosity before technology. 

  3. Learning belongs, first and foremost, to the learner. 

  4. AI suggests, people decide.

  5. We generate value, not just prompts.

  6. Share internally before sharing externally.

  7. Each pitanga tree grows at its own pace, but all in the same direction. 


Why a pitanga tree?


Pitangueiro, Pitangueira, Pitang person: that is what we call those who are part of Pitang. The metaphor wasn't invented; it was already in our name. The AI movement didn't need a name imported from the outside, because it already had its own roots. 


Like any tree, the AI Tree has five parts. Each has its function, and none can be rushed or skipped. 


Roots: people and culture 


The roots are not visible, but they decide whether the tree withstands a drought. Here, the root is culture: real autonomy, learning as personal capital, and belonging. "People before pixels" has always been more than a slogan: it is what sustains everything that grows afterward.  


With the manifesto, I also started saying “People before tokens”. Tokens are the units with which AI reads and writes, but the idea is the same as always: technology changes, and people still come first. 


Trunk: support structure and governance 


The trunk provides firmness, but it doesn't decide where the tree grows. Here, it is formed by Pitang Labs, our innovation cell dedicated to technological experimentation; by FinAI, which tracks the costs and ROI of AI usage; and by best practices in security and responsible use. Governance, for us, means providing the confidence to take risks, not pruning curiosity. 


Canopy: the community learning together 


The canopy is where the tree breathes and interacts with the environment. It is our AI Community of Practice, where Operations learns from Marketing, Support exchanges with HR, and Sales discovers what Finance has already tested. The canopy doesn't belong to any single area; it belongs to whoever participates. 


Fruit: concrete results 


The fruit is the proof that the tree is alive, but it only appears after the roots, trunk, and canopy have done their job. That is why every fruit comes with a number, even a small one. Fruit without measurement is just decoration. 


Seeds: what we share outward 


The seed is the part that leaves the tree to be planted somewhere else. That is how we view external sharing, such as this article, podcasts, and open sessions. It is not showing off: it is planting outside something that has truly grown within. 


Fear exists, and we take care of it 


I know that this change sparks enthusiasm and curiosity, but, for many people, insecurity. There is no point in pretending the fear doesn't exist. It diminishes when we learn, experiment, and share. 


That is why the manifesto is not a one-way demand. The principles themselves were designed to reassure: no one needs to grow at the same pace, learning belongs to the learner, and the final decision is always human. The company also makes commitments: approved tools for safe AI usage, continuous training, communities of practice, and room to experiment without fear of making mistakes. 


In practice, this happens on three fronts: 


  • Constant communication: every week we share with the entire company what is evolving, what teams have discovered, and what is still being tested.


  • Approachable leadership: leaders open to conversation and guidance, with space to resolve doubts and learn alongside the team. 


  • Listening: we ask people what they feel, what they fear, and what support they need. It is this picture that shows where we need to offer more support. 



This last point is one of the things I am most proud of at Pitang. We spoke with professionals in development, quality, design, project management, technical leadership, and system maintenance—from newcomers to those who have been with the company for many years. I won't name names, because what matters is the collective picture. 


What people told us 


The recurring sentiment is this: the apprehension was real, but it gradually diminished with practice, clear rules, and exchanges with colleagues. 


"What if AI breaks what already works?" 


The most frequently cited fears were losing quality, seeing the AI stray from what was requested, and breaking business rules in legacy systems. In the quality assurance area, the concern was that faster deliveries would turn the testing queue into a bottleneck. In less technical roles, the doubt was whether AI would be useful for the job. And some admitted frankly: "It's fair; I even feel afraid sometimes. But AIs are here to stay, and we need to adapt and never stop learning." 


Less typing, more deciding 


Many developers barely write code directly anymore. Today, they write context, rules, and plans, and review what the AI produces. "Our role has become much more about orchestration," summarized one developer. 


In system maintenance, investigating a bug could take hours; now AI points out the probable cause, and the person confirms it. In project management and design, the gain shows up in more complete requirements and prototypes ready to test in days. And, in quality assurance, AI brought people closer to automation without diminishing the value of manual testing: "both complement each other." 


AI makes different mistakes 


Those who review AI's work have learned to be suspicious of another kind of error: safeguards for scenarios that will never happen, features that no one asked for, business rules understood generically, and explanations that seem correct but were built on incomplete context. Because of this, understanding the business has become even more important. "I treat the AI's explanation as a starting point, not as the truth," shared one developer. 


What they say to those who are still afraid 


  • "Use AI, take advantage of everything it can offer, but never trust 100% of what it delivers." 


  • "AI doesn't diminish the developer's importance; it shifts where our value is concentrated." 


  • "Don't let AI replace interactions with people." 


  • "Don't put it off until tomorrow. Start now." 


From one week to a day and a half: one fruit among many 


This text is about culture, but an example of how it translates into results is worthwhile. In the maintenance of a legacy system, processing done item by item, triggered manually, used to take about a week. With AI helping to map bottlenecks and refine the plan, the team built batch processing: today, more than 600,000 items run on their own in about a day and a half. 


The solution took three days to develop; without AI, it would have taken weeks. And, as with every fruit of our AI Tree, AI executed, but people provided the direction. Other fruits and learnings will appear here on the blog in upcoming articles. 


What now? The question left for you 


The AI Tree is not a program with an end date. It is a tree that we nurture every day, at the pace of each area, but in the same direction. And what we have learned so far extends beyond Pitang: no tool, no matter how good, transforms a company if people do not go along with it. 


So, I leave a question for you, wherever you are: 


In your company, is AI growing with people or despite them? 


If the answer isn't clear yet, a good starting point is the same one that worked here: listen. Ask what people fear, what they have already discovered, and what support they need. And, if the question is for you as a professional, it is even simpler: which activity in your work could be improved with artificial intelligence? 


Start. Experiment. Share. Learn. Teach. Like any seed, this text exists to plant something elsewhere. 


Here, AI doesn't grow haphazardly. Here, we nurture, learn, and grow with it. 


Do you want AI to grow with people in your company too? 


Pitang helps organizations adopt artificial intelligence in a structured way, with the same care we have internally: listening to people, creating support so that everyone learns, and measuring results.


Want to learn more?

Schedule a conversation

Recife

Cais do Apolo Street, 222 - 13th floor

ZIP: 50030-220 - Old Recife Recife/PE

Tel: +55(81) 3134 5200

Fax: +55(81) 3134 5374

São Paulo

Joaquim Floriano Street, 466 - Suite 1001 - 0

Brascan Century Building

Corporate, Itaim Bibi

São Paulo/SP

Recife

Cais do Apolo Street, 222 - 13th floor

ZIP: 50030-220 - Old Recife Recife/PE

Tel: +55(81) 3134 5200

Fax: +55(81) 3134 5374

São Paulo

Joaquim Floriano Street, 466 - Suite 1001 - 0

Brascan Century Building

Corporate, Itaim Bibi

São Paulo/SP

Recife

Cais do Apolo Street, 222 - 13th floor

ZIP: 50030-220 - Old Recife Recife/PE

Tel: +55(81) 3134 5200

Fax: +55(81) 3134 5374

São Paulo

Joaquim Floriano Street, 466 - Suite 1001 - 0

Brascan Century Building

Corporate, Itaim Bibi

São Paulo/SP