The true transformation of AI: from technology to people
The true transformation brought about by AI is not technological, but human — it concerns roles, culture and, above all, the speed at which we are able to relearn.
A few months ago, I wrote here about the responsibility that falls on those of us who build digital products in the age of artificial intelligence. I went back to that piece this week and realised that the question I was asking myself back then has changed. When we first started talking about the AI SDLC — that is, how to build digital products using AI itself — the question was why we should do it. Today, another question, one that is both more uncomfortable and more exciting, is on my mind: how do we change ourselves to keep pace with what AI already enables us to do?
I think about this often, and I always come to the same conclusion. Technology is no longer the bottleneck. Now it’s us, the people. The new limit isn’t what the machine can do, but the speed at which we learn and transform ourselves. I suppose it’s a recurring question whenever there’s a technological transformation.
The shift in role: from executors to decision-makers in the age of AI
I must confess that for a while I viewed this change with some unease. If AI generates, what is left for us?
The answer, as I have come to understand it, is precisely the opposite of what I feared: AI does not dilute our responsibility; it elevates it. We cease to be executors who generate output and become something more demanding — product builders who provide context, review what has been generated and, above all, make decisions.
Our role is no longer to produce every single piece, but to be at the heart of the most important decision points. That is why what we call ‘human in the loop’ does not strike me as an option, but as a necessity: we cannot leave AI to its own devices, because it makes mistakes, it goes off on a tangent, it generates more than you ask for, it can be complacent…
It makes proposals; the judgement, the intention and the validation remain ours. Every day I find more meaning in talking about ‘product sense’, ‘design sense’ or ‘engineering sense’: that judgement which is difficult to codify is, precisely, what the machine lacks.
And this change cannot be improvised on a team-by-team basis. It requires a roll-out plan designed for the whole company, because the hardest part isn’t the tool — that can be bought or learnt — but the cultural shift and the new way of working.
Our responsibility, I believe, doesn’t end with us learning: we have to help our organisations — in my case, Telefónica — support people through this transformation. No one should be left behind simply because they haven’t had time to relearn. That’s how I see it personally.
Accelerated innovation: how AI turns discovery into a competitive advantage
There’s one aspect of all this that really excites me. We have more resources than ever to carry out discovery, and that changes the very nature of innovation.
Implementing continuous discovery workflows and accelerating our funnels — forgive the Anglicisms — means something very specific: testing a hypothesis has gone from being expensive and slow to being fast and cheap. We can see sooner what’s worth pursuing and what isn’t, discard what doesn’t work early on, and focus investment on what does.
Innovation is no longer a shot in the dark; it has become what it always should have been: a process of continuous learning.
From idea to production: speed, efficiency and new risks
And once we know what to build, the rest of the journey is also shortened. We can write precise specifications in a fraction of the time, code them more quickly and — this strikes me as just as important — ensure they comply with legal requirements, security standards and QA, and deploy the code to production at a speed that previously seemed impossible.
But it is not just the speed that is valuable. It is what that speed gives us in return: time to think more clearly, prioritise wisely and take care over what we deliver.
I find this extremely important because, from experience, it’s very easy to fall into the trap of ‘doing more just for the sake of doing more’. Having greater capacity without focus can lead to wasting resources, creating noise or getting stuck in an unproductive loop. Here, more than ever, that ‘sense’ we were talking about earlier is key once again.
Generative AI and agents: from content to autonomous execution
It’s important not to confuse the two levels. Generative AI gave us the ability to produce content based on an intention.
Agents go further: they chain together tasks, execute complete workflows and coordinate processes to the point of approaching production.
But the more capable they are, the more important it is who governs them. An agent without human supervision isn’t autonomy; it’s delegated risk.
The key question is no longer what they can do, but what we decide they should do, within what context and under what control.
The customer as the sole measure of value in the age of AI
As I write all this, I cannot help but think that none of it makes sense if it does not reach the customer.
It is easy to become obsessed with internal efficiency, but quality is not measured by what we produce, but by the customer’s trust.
AI used well means better experiences, more useful products, less friction and greater foresight — always whilst respecting privacy and trust.
The true indicator isn’t on our dashboards: it’s whether the customer returns.
Real risks of AI: cost, control and governance
It would be naive to ignore the risks.
There is an economic reality (every token counts), a technical one (AI makes mistakes) and another, more relevant one: governance.
How we scale whilst maintaining quality, how we avoid dependencies, how we protect talent and how we guarantee privacy and security are not problems of tools, but of judgement and strategy.
“Context, context, context”: it’s not about asking for more, but about asking better.
The start of a more profound transformation at Telefónica
None of this is an isolated experiment. It is the start of a more profound transformation at Telefónica.
A transformation that forces us to ask ourselves: are we learning quickly enough? Are we transforming people or just the tools? Does what we build genuinely improve the customer’s life?
We do not have all the answers, but one thing is certain: these are the right questions.
Our value proposition: technological leadership with real impact
In this context, Telefónica is taking on the challenge of becoming the best gateway for citizens to digital technologies, driving a transformation based on five key pillars: consolidating its European leadership by contributing to technological sovereignty; building a more innovative and competitive company through simplification, efficiency and talent; offering more and better services supported by the best network and customer trust;
moving forward with ambitious and rigorous management; and reinforcing its role as an institutional benchmark in Europe. All of this with a clear objective: to transform artificial intelligence into real and sustainable impact for the business, customers and society.
