As the year comes to a close, it’s a time for reflection—a moment to revisit ideas that have matured and to spark new ones. Rather than offering a retrospective or a list of trends, I’m sharing insights on AI, inspired by conversations with Emmanuel Vivier.
It’s a starting point—for you to comment, expand, and explore.
Above all, in this season, let us hold on to what truly matters. Beyond possessions, it is family, friends, and the shared moments—fragile and fleeting—that are life’s most precious gifts.
So instead of wishing you happy holidays—a brief and often conventional joy—I wish you holidays that truly resonate with your soul.
Eureka!
AI shouldn’t just answer our questions—it should cultivate in us the art of asking them. It should awaken our critical thinking, especially when confronting its biases. To question is to challenge the obvious and reject passivity in the face of pre-packaged knowledge.From Attention to Appreciation
In an era of AI-driven automated creation, true rarity lies in the meaning behind content and in our ability to discern what truly deserves to be seen, read, or experienced. The culture of abundance calls for an ethic of the eye—a deliberate, mindful way of seeing that values depth over distractionThe False Debate of 100% Human
Who wonders if the paintbrush creates instead of the artist? Tools, whether modest or powerful, have never replaced the artist—they reveal their vision. AI is no exception. Neutral by nature, it can imitate, reproduce, and organize, but it becomes transformative only in the hands of those who know how to wield it. The real question is not about the tool—it’s about who holds the brush.
AI, a Revealing Mirror
It’s often said that those who use AI will surpass those who don’t. But the reality is more nuanced: AI amplifies what already exists. It reveals brilliance just as much as it accelerates mediocrity. While it propels the world forward, those who choose to ignore it risk inevitably falling behind.
AI Doesn’t Know How to Be
AI doesn’t feel the impatience that breaks a cycle, the weariness that compels us to turn the page, the thrill of a sudden intuition, or the irritation that exposes a deeper flaw within us.
It doesn’t sense the weight of silence, the spark of a glance, or the subtle fear that accompanies hesitation. It doesn’t understand the spontaneity that enlivens a conversation, the ambiguity that confuses, or the vulnerability that reveals who we are.
And yet, aren’t we at risk of ascribing these virtues to it—virtues it doesn’t possess—simply because we stop fully exercising our own humanity? Or worse, could the machine eventually judge us—too slow, too flawed—to keep up with its relentless pace?
The Era of Outsmarting?
We’re being sold the “agentic” era: hyper-powerful AI agents, universal tools to do everything and amplify anything.
But the real challenge lies elsewhere. If AI mimics intelligence, humans know how to be clever, witty, and resourceful.
So, what happens when an army of bots floods in? Humans fight back. They create new bots to outsmart the first wave. Take Daisy, for example: an AI disguised as a clumsy grandmother that has already scammed phone scammers.
In the end, if machines aim to think, humans will teach them to think twice.
When AI Takes Action
Large Language Models (LLMs) have revolutionized text creation, but they are now evolving into tools that are more practical and action-oriented. Enter LAMs, capable of transforming instructions into actions, and CALM, developed by Google DeepMind, which pushes the boundaries even further by opening up new possibilities.
With CALM, multiple specialized AIs can collaborate seamlessly without requiring complete reprogramming. For example, a general-purpose model can team up with an expert model to perform complex tasks, such as writing code or solving logical problems. This approach relies on a lightweight mechanism of interaction between models, preserving their individual strengths while enhancing their collective capabilities.
And this is just the beginning. AI will continue to surpass these innovations, becoming increasingly adaptive and action-oriented. This rapid evolution calls for urgent reflection on governance, values, and human autonomy in a world where AI is not just operational but also decisional.
Perfectly Imperfect
In the face of AI that standardizes perfection, human traces—mistakes, smudges, scribbles—will become marks of authenticity and emotion. Imperfection will be a luxury. Uniqueness, a rare privilege.
Shaken Foundations
Titles, topics, and keywords—these “handles” that guide our access to information—are being redefined.
A recent article highlights their central role: shaping expectations, filtering choices, and even influencing creation. But with AI, these markers are now generated automatically from analyzed content.
When opaque systems shape our expectations, what happens to independent judgment? For creators, losing control of handles risks distorting their message. For consumers, delegating these cues to AI threatens curiosity and critical thinking.
Handles will persist, but their nature and influence will be profoundly transformed.
Hallucinations-as-a-Service
AI hallucinations will become monetizable in unexpected industries.
The Hidden Costs of AI
AI comes with invisible costs: energy consumption, obsolescence, maintenance, and complexity that can be difficult to manage. Sometimes, the most sustainable solution remains... human.The Emotional Challenge
AI must understand human emotions—not to mimic them, but to enrich and adapt its interactions with us.Selfpressionism
Generative AI, capable of decoding and reinterpreting deeply human content, combined with quantum computing and advancements in neuroscience, opens dizzying new possibilities. By exploring our thoughts, emotions, and creations, it doesn’t merely produce—it illuminates who we are.And if fact, we may discover that the true mystery isn’t the machine—it’s humanity itself.
Outskills
The future doesn’t lie in the tool itself but in how it becomes an extension of our uniqueness. It depends on our ability to delegate to AI tasks that don’t require our essence, avoiding the risk of losing ourselves in abundance. This won’t be an era of mass content but of scarcity, where each custom agent embodies a unique vision and nurtures what sets us apart.
Monetiz-AI-tion
Sam Altman has made a bold bet: the emergence of the first “one-person unicorn”—a startup valued at $1 billion with just one employee, thanks to AI. A fascinating vision, but one that raises a far more troubling question: can such a future be built without first creating a society capable of embracing it? And can a “winner takes it all” world, pushed to the extreme, ever truly be sustainable—or will it simply become unbearable?
Ethics or AIthics?
Behind this paronomasia lies the real challenge of the coming years in AI. Ethics, as universal principles guiding our decisions, or AIthics, a new form of morality shaped by artificial intelligence, where algorithms redefine what is just and desirable.
The real question: will we choose to be the authors of these values, or merely spectators to a logic imposed by machines?
The Genius Flaw
A (super) AI that is right 9 times out of 10... but hallucinates the 10th. Progress or peril?
UberAI
How do we collaborate with an AI that could surpass human intelligence? If intelligence reflects human expectations, what safeguards can we establish to ensure it remains aligned with our values? More importantly, how do we guarantee that it serves humanity without harming its well-being?
Cyborgs, Transhumans, AI Assistants
What does the future hold for the alliance between humans and AI? The real challenge isn’t technological—it’s existential.
Underground AGI
If a government were to develop an AGI (Artificial General Intelligence)—an AI capable of matching human intelligence across all domains—would it choose transparency, announcing it to the world, or exploit it in secret, much like the early nuclear weapons, to assert dominance?
MD
Good insight 😌. Can i translate part of this article into Spanish with links to you and a description of your newsletter?
So good! Number 10 is happening already over at Springboards AI. Number 14–this is the mindset shift we need