Interview: Ricardo Chaves on AI, Leadership, and the Human Lessons Behind Technology
Ricardo Chaves reflects on AI, leadership, and how history, ethics, and adaptability shape successful transformation in today's world
It’s not often you get to sit down and formalize a conversation you’ve been having, in bits and pieces, over nearly a decade. For me, today’s interview feels exactly like that. I’ve known Ricardo Chaves since long before AI became the force reshaping industries it is today. We’ve crossed paths in different roles, collaborated on diverse projects, and shared more than a few late-night discussions—whether about the future of business, technology trends, or sometimes just history and literature, two of Ricardo’s enduring passions.
What has always struck me about Ricardo is how seamlessly he moves between worlds—starting his career in Law, diving into public sector finance, scaling businesses from scratch, leading strategic transformations in banking and payments, and now standing at the forefront of AI-driven business reinvention. He’s a rare blend of conceptual thinkers, pragmatic leaders, and lifelong learners who refuse to stay confined to a single box.
Given Building Creative Machines' focus on generative AI and the new creative paradigms it unlocks, I couldn’t think of a better guest to unpack not just the tech but also the broader human, ethical, and leadership dimensions that come with it. Ricardo’s career arc, perspective on history’s lessons, and hands-on experience navigating technological revolutions give him a unique vantage point, and I’m thrilled to bring that perspective here.
Credits: BPI Artificial Intelligence Center of Excellence, founded by Ricardo Chaves
Your career spans a fascinating journey — from a foundation in Law to strategic consultancy, leadership in the industrial sector, and nearly a decade shaping the Fintech and Banking landscape. How has this multidisciplinary path influenced your perspective on decision-making in an era increasingly driven by data and AI?
Starting my career in the legal field, such a different one from those where I spent most of my professional life (strategic consultancy, business transformation and AI industrialization), I believe boosted my curiosity and humility to a very high degree. It translated into openness to new ways of thinking and hard work in order to master very different pieces of knowledge. From the legal background I took a strong conceptual thinking, from strategic consultancy an ability to grasp a problem through an end to end framework, from transformation leadership the focus on holistic planning and change management and from AI industrialization a disruption and scientific mindset. Answering your question, transforming successfully with AI a business organization, especially an incumbent (as opposed to a digital native), demands a very broad set of skills and perspectives: understanding the models’ math (its strengths and weaknesses), aligning business processes, sometimes radically, to an “AI first” operating model, master technology to contribute on how to change the delivery model and being an experienced leader in supporting a complex change journey. Changing paths strengthened my adaptability. I never felt incapable but also never felt totally ready: this fueled my constant learning path.
As someone deeply passionate about History and Literature, do you believe these disciplines offer lessons that can inform how we develop or govern Artificial Intelligence systems? Can the human narratives and ethical dilemmas in these fields counterbalance to purely data-driven approaches?
It is said that Umberto Eco commented that those who don’t read live only one life while those who read live countless lives. The key idea here is that reading gives you perspective. Literature and its thousands of characters and plots teach us that things rarely follow a steady path. There is randomness, there is agency, there is human diversity, and there is society and its dynamics that can so often go the right or the wrong way. Another sentence I heard that is also helpful is that history is not the story of the people who were dumber than us but of the people who lived before us. This is looking at past experiences as a way to gain more perspective in our short lives. When we discuss ethics in AI, we are fundamentally trying to sort out right from wrong in an area where innovation, curiosity and ambition are pushing the AI frontier in terms of capabilities and uses wider every day, alongside the fact that such a disruptive technology can have impacts that we don’t foresee in advance. Gutenberg invented the press to sell catholic bibles, but the press stood as one of the main levers for protestant reform. The nuclear chain reaction was invented out of mere scientific curiosity, but it gave rise to the most powerful weapons mankind has ever produced, which can destroy Earth entirely. There is no reason not to take these lessons into consideration when we look at a technology that, despite all its potential for good, can have similar global, uncontrolled or dire effects on mankind. This means that responsibility must take place alongside the discovery process, making sure that the development is aligned with human interests. Of course, when innovation is at stake, this cannot mean that we only allow the development of totally harmless evolutions because, sometimes, it’s impossible to anticipate wrong usage in advance. When cars were invented, no one had in mind car crashes. When traffic became more complex, rules were invented, and later, security measures became generalized. I’m an optimist and I think AI will transform the way we live in crucial areas: education, health, production, science, etc. But we shall never believe we are invincible. We shall remain cautious and ready to refrain whenever the alignment is at stake or the unknowns are such that it’s better not to take risks.
You’ve witnessed the evolution from early analytics and ‘traditional AI’ to today's surge in Generative AI. In what ways does this current hype cycle differ from the transformations you've observed over the past two decades, and where do you see potential overstatement versus genuine disruption?
An analysis of the last 25 years in AI has to be split into two different realities: disruption led by attackers and disruption within the incumbents’ spectrum. If we look at what attackers such as Amazon, Spotify, Netflix, Meta, Tencent, etc, achieved in the sectors they disrupted, it’s undeniable that AI, particularly the “traditional AI”, such as predictive Machine Learning, predictive Deep Learning, Recommendation Engines, or any “mix” involving these techniques, was already so powerful that allowed them to achieve an unique level of performance, evidenced on how these players outperformed their competitors and reshaped their markets, with new and better solutions that their incumbent competitors were totally unable to cope with. Amazon had such a radical belief in the power of “traditional” predictive AI that it focused all efforts on automating every piece of its e-commerce planning, sales strategy and delivery model. If you look at the slow pace of AI adoption of most giant traditional retailers, what I take from it is not that traditional AI wasn’t ready to disrupt, but that it is really difficult to make an incumbent enter into a radical change process, even when survival or big growth opportunities are at stake. So, the hype generated 15 years ago was totally justified, given what it allowed to accomplish, but it was also very hard to implement in traditional companies with dominant market positions. There are two factors, though, that make this later hype, driven by Generative AI, different from the previous one. The first is the fact that User Experience and the type of use cases Gen AI delivers are much easier to understand by non-technological profiles. When someone witnesses a high-quality bot with a human-like voice and high accuracy, for instance, in a call centre, it is natural that the “belief” that AI can really transform the business is more evident. The second is the fact that these models have shown that their general capabilities, with some expert fine-tuning or minor adaptation, could make them fit to solve problems that range from image recognition, conversation, protein structures, and reasoning, among so many others that are showing up every day anew. This is to say that we are closer than ever to a general level of intelligence, at least for the level that is required for practical business purposes. In a nutshell, as difficult as the change may be for the ones that start from a traditional baseline, disruption is clearly the word to describe what the current AI surge will do to everything we know.
Credits: Jornal Económico, 2024, during a public speech by Ricardo Chaves
Having led analytics initiatives across multiple industries, particularly in Payments and Banking, how do you assess the tangible impact of AI and advanced analytics on business outcomes? Are there common pitfalls or misconceptions that persist across sectors?
My experience taught me that the transformation required to achieve impact depends on 4 factors: infrastructure and data readiness, industrialization of analytical assets, technological delivery model and business change. But prepare for the worst news: the bigger the impact, the harder the adoption funnel. Business change is by far the harder nut to crack, and most of the failures lie here. The first advice I would give is to, in the early stages, focus on laying the foundations right, right-sized for the transformation ambition. Invest in infrastructure, build a comprehensive team and prioritize technological investments that enable the deployment of AI assets. The second is to focus on an ambitious portfolio so that AI becomes relevant at a strategic level, involving all the key areas and touching the major operations of the business. At this stage, measuring the pace of adoption rather than the business incomes is better to ensure implementation focus and pace. Change is hard, and most of the business impact can be delayed until the organization adapts to the “new order”. Demanding results too early may discourage the transformation leaders and give strength to the “old school” opponents. Third, once the adoption is halfway through, review the performance management metrics focusing on real business impact, adjust to the new “AI first” model and bring discipline towards a continuous improvement mindset.
Reflecting on the projects and leadership roles you’ve held in the past, how do you think access to today's generative technologies and AI capabilities would have reshaped the outcomes or the approach you took at the time? Looking forward, how do you envision the long-term trajectory of these technologies altering the landscape of industry and leadership?
I tend to see the generative AI wave as complementary to the previous one. Predictive and traditional AI are still at the core of a lot of business process transformation: proactive sales management, hyper-personalization, churn management, dynamic pricing, predictive maintenance, demand forecasting, etc. Also, I think that tough Gen AI appears to be easier to implement at the corporate level, that’s not exactly the case. Setting up a personal AI assistant for personal usage with the latest Chat GPT model seems quite easy for everyone, but setting up a bot to automatize the Help Desk that is able to manage the bank’s whole knowledge base, from product forms to pricing, to bring the relevant client context, and be able to answer with minimum or none bias, hallucination or error, is the same as comparing driving your car fast in a highway to being ready to compete in a Dakar race. I believe that the organizations that have invested early in setting up the foundations for traditional AI will now be able to take fuller advantage of the gen AI surge. Having said this, generative AI shall have a remarkable trajectory of its own: it will probably change significantly the way most jobs are done today, and its “general” capabilities will continue to open new use cases where it can add value in ways we can’t even foresee today. There are already some challenges specific to Gen AI, such as the capture of its benefits in terms of productivity by the employer: people use it but at a personal level, not sharing it as a new way of work, keeping its impact away from the company’s P&L. I believe the right way to overcome this is to embrace it even further: promote those that innovate and share, reward productivity enhancements sharing the profit with the change leaders and use Gen AI to make AI fun, down to earth and a constant growth opportunity for everyone.
Credits: DSPA Insights, 2023, Data Science event featuring Ricardo Chaves
Talking with Ricardo Chaves always feels like recalibrating your compass. He reminds you that no transformation—whether powered by AI, data, or human ambition—happens in a vacuum. It’s about people, systems, and the narratives we build around them. What I appreciated most about this exchange is how it felt like an extension of conversations we've had over coffee, after meetings, or between life’s busier moments, now framed against the backdrop of one of the most pivotal technological shifts: the rise of generative AI.
It’s easy to get caught up in hype cycles or to focus narrowly on tools and models. Ricardo brings us back to fundamentals—curiosity, responsibility, and adaptability—and how these apply whether you're leading an incumbent through AI industrialization or trying to understand where creativity and machine intelligence intersect.
I am grateful as always for his thoughtful take and equally excited to continue this conversation—because with Ricardo, there’s always something new to learn and the next frontier to explore.
If you enjoyed this interview, explore dozens more exclusive conversations with leaders from Anthropic, Amazon, Google, Harvard, TikTok, and beyond—each shaping the world by Building Creative Machines.