Fork Addiction

Picture tomorrow morning’s front-page headlines:

Seasonal fruit, fresh vegetables, or dripping burgers: the average person spends over three hours a day eating. Shock investigation into the scourge of “fork addiction.”

Or even:

History podcasts, symphonic music, conspiracy debates: we spend five hours a day glued to our speakers and headphones. Is it time to ban vibrating membranes?

Absurd? Of course it is.

Yet that is precisely the depth of analysis we are treated to whenever digital technology is discussed. For years, mainstream media, pundits, and politicians have brandished the exact same bogeyman: “screen addiction.”

A hollow, convenient, and profoundly depoliticized buzzword.

Confusing the Vessel with the Content

The first flaw is technical and semantic: blaming the tool instead of the activity makes no sense.

A screen is nothing more than a photon receiver, a backlit panel—just as a transistor or Wi-Fi are merely physical conduits for information. Claiming someone is “hooked on screens” makes as much sense as diagnosing an “addiction to paper” on the grounds that we indifferently print sci-fi novels, tax bills, poetry, grocery flyers, or far-right propaganda tracts on it.

Lumping every single use under the catch-all banner of “screen time” conveniently erases all nuance. In those panic-inducing statistics of multiple hours a day, what are we actually tallying?

  • Work hours (where people are explicitly required to sit in front of a monitor for eight hours a day);
  • Everyday commerce: booking train tickets, ordering a book, click-and-collect grocery shopping, comparing energy quotes, paying for a gym membership;
  • Paperless administrative procedures;
  • Private conversations with family, friends, communities, colleagues, or comrades;
  • Creating, learning, reading;
  • Consuming cultural content, news, and entertainment;
  • And, finally, passively scrolling algorithmic feeds.

Lumping together an hour of coding, an hour on video call with your grandmother, and an hour of mindless doomscrolling on TikTok as one single pathological behavior is pure intellectual laziness.

And frankly, nobody is truly fooled. Whenever people are asked about “screen addiction,” they invariably bring up video games and social networks. They never talk about the hours they spend reading articles, watching movies, or following sewing tutorials on YouTube.

Addiction is a Serious Matter—Let’s Treat It Seriously

Let’s be clear: I am not dismissing the reality of psychological distress or the mechanics of dependency. Addressing addiction is vital. Real people—both young and older—are losing sleep, becoming isolated, and suffering.

But addiction belongs to the medical and psychological realm. It demands precise clinical definitions, identifiable symptoms, and appropriate therapeutic support.

When someone experiences distress tied to their digital habits, their need for help is entirely legitimate. However, slapping the label “screen addict” onto that distress conceals the true source of the problem. It fixates on a crude, visible symptom rather than investigating the root causes.

And that is entirely by design.

The Attention Economy and Predatory Design: How the Trap Works

What we are up against is not some fetishistic obsession with a small, glowing rectangle (unless we’re talking about my love for the Triforce, but that’s another story). We are up against digital products engineered with surgical precision to capture and hold our attention for profit.

Screens do not make people addicted. What does is:

  • Recommendation algorithms, fine-tuned to trigger outrage, anxiety, or variable intermittent rewards (the classic “slot machine” effect).
  • Predatory interfaces riddled with dark patterns: infinite scroll, autoplay, deceptive push notifications, and worse.
  • The advertising business model, which converts every single second of our available brain time into monetizable ad inventory.

When you lose 45 minutes without realizing it, watching an endless loop of absurd videos, it isn’t some mysterious failure of willpower in the face of a black mirror: it is the direct outcome of entire teams of engineers and designers paid top dollar to exploit our cognitive biases.

And when researchers blow the whistle and rigorously document these predatory practices, what do the platforms do? They unleash armies of lobbyists to discredit independent science, pit academics against one another, and bury inconvenient findings1.

Who Benefits from the Fog?

Why this stubborn refusal to call things by their name? Why this refusal to dissect what actually fills those hours?

Because scrutinizing usage means asking a political question: who owns the platforms whose algorithmic editorial choices shape our information diet and, by extension, our lives? Billionaires locked in an arms race of lobbying to guarantee nothing threatens their bottom line.

So instead, we talk about “screen addiction.” We dump the entire burden of responsibility onto the individual (or onto supposedly negligent parents). We moralize, we guilt-trip, we hand out tips for “digital detoxes” and app screen-time limits.

Meanwhile, we conveniently avoid the core issues:

  • Why do we allow predatory design mechanics to thrive without strict regulation?
  • Why aren’t we challenging the economic model of surveillance advertising?
  • Why aren’t amplification algorithms regulated the same way we enforce food safety and hygiene standards?

If your food is laced with addictive or toxic chemicals, your fork isn’t the problem. The problem is the manufacturer poisoning the plate and the regulator letting them get away with it. It is high time we dropped the moral panic over screens and started collectively, politically regulating what goes onto them.

  1. “A $17.1 Billion Vote of Confidence That Jonathan Haidt Was Right”, Emma Goldberg for The New York Times, August 27, 2026. ↩

Algorithms of Isolation

As I write, Apple and Meta are rolling out their latest gadgets: intrusive glasses that allow anyone to film you without your knowledge, and a watch with a permanent listening mode, recording all your conversations and interactions with your environment.

They use the same arguments. They promise convenience, efficiency. No need to recall every detail of a conversation—it’s transcribed in real time. No need to remember your schedule, appointments, tasks, or ideas. Everything is recorded, analyzed, and summarized for you. No need to think, no need to remember. You won’t make mistakes anymore because you won’t make any effort.

But I don’t think this is about progress. I believe it’s about control.

Contrary to what the tech giants claim, these devices are not tools to enrich life; they are replacements for it.

Why bother listening to someone’s words, understanding their identity, the way they express themselves, their hesitations, their contradictions… when an AI can summarize it all for you? Why reflect on how you emotionally engage in a conversation, how you mobilize your attention and memories to understand the other person, when an AI can do it for you?

Why engage in the messy, unpredictable, and emotionally challenging yet rewarding work of human connection when you can retreat into a bubble?

In my view, that’s not empowerment. That’s surrender.

I read Shoshana Zuboff a few years ago, and I remember her stance, which at the time seemed quite science-fiction-like to many: algorithms do not merely know us; they shape us.

With this reorientation from knowledge to power, it is no longer enough to automate information flows about us; the goal is to automate us. […] do not merely know our behavior but also in different ways try to modify and shape that behavior toward profitable outcomes.
Shoshana Zuboff, “The Age of Surveillance Capitalism”.

Even today, we collectively struggle to understand what this means. Just look at the debates about banning social media for children under 15—it reveals how little we’ve considered the implications of algorithms. We ask whether children are mature enough to use these tools, but we don’t ask whether these tools are mature enough to be used by children. In fact, we don’t even ask whether these tools are mature enough to be used by adults…

We all know very well what these tools create: isolation. Loneliness. Passivity. Docility. Predictability.

Paul Klotz’s Solitude Capitalism (a concept I discovered thanks to Tristan Nitot) frames this perfectly. We are not just consuming technology; we are collaborating in our own isolation.

The more isolated we are, the more predictable we become. The more predictable we are, the easier it is to sell us things, ideas, products, even versions of reality. The more isolated we are, the easier it is to sell us services that keep us in that isolation, including by creating an impenetrable interface between us and the real world.

I define “Solitude Capitalism” as an economic system in which loneliness has shifted from being an unintentional phenomenon to a resource that is deliberately sought out and exploited.
Paul Klotz, Interviewed by Le Monde on August 26, 2026

This is not innovation. This is not progress.

Our hegemonic culture is one of performance. We must perform everything: our studies, our skills, our relationships, our parenting, the way we exist for others, the way we perceive ourselves. We have tools to measure, quantify, and optimize everything. But to what end? Consumption.

It’s not about maximizing our well-being, but our subjugation to a system where consumption is the only goal. These gadgets are performative: they describe a hostile world to be dominated by technology. They warn us against the chaos of the outside world and promise to act as a secure, protective interface. And this can only work if we adhere to the idea (I first wrote “buy into the idea”, and that says a lot) that the world is fundamentally unfit for our needs, and that we must consume products and services to protect ourselves from it.

It’s a vicious cycle: the more we consume, the more isolated we become, and the more isolated we are, the more we consume.

Everything is designed so that no one reminds us that what is lost is not just serendipity, privacy, or the ability to focus: it is the ability to think critically, to engage deeply, to resist, to create, to feel.

The legacies we fail to pass on are not just material. They are the very things that make us human: effort, friction, the unpredictable beauty of real connection, disorder, boredom, suffering, but also the joy of being alive.

Rethinking Travel

10 days ago, I went to Berlin for a company event. I asked to go by train, instead of by plane. I spent 26 hours traveling by train from Périgueux to Bordeaux, then Paris, and finally Berlin, returning through Mannheim.

The journey was long but beautiful, with stunning landscapes. I enjoyed exchanging anecdotes with my neighbor, canevasing, working on my laptop, taking breaks to stretch, or lending my Switch to colleagues for the biggest Mario Kart tournament ever.

Without the train, my alternatives would have been1:

  • 🚗 Driving from Périgueux to Bordeaux: 60 Kg of CO2 (half if carpooling)
  • ✈️ Flying from Bordeaux to Berlin: 500 Kg of CO2

By train, I only emitted:

  • 🚃 Périgueux to Bordeaux by TER: 8 Kg
  • 🚆 Bordeaux to Berlin by TGV and a bit of metro: 8 Kg of CO2
  • and as I had to add two hotel nights: 20 Kg of CO2

But that’s still about 15 times less!

We can’t always avoid flying2, but when it’s possible, I consider my duty to think of alternatives, especially since I had the time and resources to do it.

Un montage de différents selfies pris en chemin.

  1. The estimates come from the Ademe carbon footprint calculator ↩

  2. I have family in the French West Indies, and although I haven’t visited in nearly a decade, I plan to make the trip soon. Unfortunately, there aren’t many options for quick back-and-forth travel that take less than three weeks! ↩

Weapons, GenAI

When one side develops a weapon, it is traditional for its opponents to seek to seize it, or to design an equally destructive weapon, in order to establish a balance of deterrence. The digital world is no exception: for every data-gathering or attack capability via increasingly sophisticated vectors present on one side, the other side develops countermeasures and builds its own modes of attack.

The emergence of generative AI in the public sphere continues the astroturfing1 that has been going on for decades, helping to erode the very concept of truth, both as reality and as possibility. If everything becomes credible, then nothing is truly true. Attempts at regulation are emerging, but they are moving slowly and proving complex, as not all players agree. Some have already undermined the truth in their own countries: destroying the truth in others is just a continuation of this process.

So, instead, a business is formed. We congratulate ourselves on having invented tools so powerful that they become ever more destructive. Efforts are made to dismantle the weak regulatory barriers already in place or about to be introduced. We praise the economic multiplier of this technology, without any in-depth knowledge of the matter.

And as always, we turn the vocabulary on its head: regulation is a prison; state aid is a prerequisite for success (we’d like to see the same logic culturally applied to social aid); moderation is a form of censorship; freedom is domination by the powerful; competition, including winning military contracts, is peace.


Having said that, what’s next? What issues would I have liked to see emerge from the French AI summit (and I’m not saying they weren’t addressed there, just that they weren’t publicized)?

I don’t think we can imagine a world today without generative AI. The technology has infused so many practices in such a short period of time… it seems impossible. But we can ask ourselves, collectively, what kind of AI we want.

From an ecological point of view, first of all, I think that the European Union’s Corporate Sustainability Report Directive (CSRD) and, more generally, the Science-Based Targets initiative (SBTi) are going in the right direction by creating a non-financial accounting of greenhouse gas (GHG) emissions.

This can only work if regulation is maintained, strengthened, and publicized over time. Knowing that a company like Microsoft exceeds its targets by 29% is important. If this overrun has impacts on the sustainability report of their clients, then only do we create a competitive situation around ecological indicators for the benefit of the planet.

But the question goes far beyond the ecological issue: we need companies to discuss and set up very precise screens on the AIs they use, and the use cases where they apply them. These screens must contain:

  • Specific use cases, and their success criteria
  • A survey of available models, their capabilities and biases
  • Test protocols with samples of use cases and an evaluation of feedback
  • An explicability framework to ensure the transparency of the decision-making process, independently audited to validate its legality and the organization’s accountability
  • A continuous monitoring mechanism, with a disaster recovery plan if the AI needs to be disconnected

I’m convinced that all these needs will create jobs in the future, from the AI choice consulting firm to the structure specializing in continuous monitoring, via the independent audit firm, or even one specializing in the specific task of the business need (you don’t audit generative AI for translation, labeling or general discussion in the same way).

The speed at which we make progress on these issues will depend on our political courage and our ability to stand up to the economic interests of the major AI players, who don’t really want this kind of competition.

It’s up to us to force them.

  1. a deceptive propaganda practice in which organizations or individuals create the illusion of widespread popular support for their product, service or point of view. ↩

Emmanuel Macron wants to see the left die

A few hours ago, Emmanuel Macron dissolved the National Assembly following the French result in the European elections11. A part of me doesn’t understand this decision, and the other is afraid of understanding it too well…

No power at European level

Let’s start with the reality of the impact: the far-right group in the European Parliament is ridiculous.

Even with 30 more seats, the Identity group remains very small (barely 8% of the seats), far behind the 180+ EPPs (European Populist Party) and the 130 Socialists and Democrats (S&D).

The activity of RN elected representatives is negligible, with very few proposals. Jordan Bardella signed 21 amendments between 2019 and 2024, compared with several hundred, if not thousands, for Manon Aubry or François-Xavier Bellamy. He is not the rapporteur for any text, compared with three for the ecologist MEP Marie Toussaint.

It’s not just the RN’s fault: the other groups are applying a kind of sanitary cordon and blocking the identity group’s room for manoeuvre (and in my opinion, that’s a good thing). The RN politician is making his presence felt to avoid a new trial for fictitious employment, Marine Le Pen having just reimbursed €300k to the Parliament…

La montée de l’extrême-droite au niveau européen est un sujet de débat mais c’est une montée lente, par un raz-de-marée.

The rise of the far right at European level is a subject of debate, but it is a slow rise, not a tidal wave.

So why is Macron acting as he is? What is the urgency?

Strengthening a two-party system

Creating a republican shield to counter the far right is not the same as creating a situation of opposition to the far right in order to create a republican front revolving around you.

First of all, when you want to create a bastion against a political opponent, you don’t borrow half his proposals. You don’t propose legislation on the basis of a majority that includes the RN.

But on a deadline like a dissolution, if you want to build a wall, you give it the means to exist, including by giving it time to be built.

By choosing such a short deadline for the dissolution (the vote will take place in 20 days’ time, i.e. the minimum amount of time possible), Emmanuel Macron is banking on the inability of the left to reach agreement, at a time when the opposition parties are exhausted by a difficult European election campaign and have neither the time nor the energy to establish programmatic agreements on national issues. Besides, the coffers are empty: even the parties that obtained more than 3% in the European elections will only be reimbursed in a few weeks, or even months.

We are therefore heading for a second round in which, in many constituencies, the presidential party will face the Rassemblement National, and Emmanuel Macron knows that in this configuration, all democrats will do their duty and “make block”.

In constituencies where an RN candidate will face someone from a party other than Renaissance, voting instructions will be given on a case-by-case basis, depending on the person’s alleged compatibility with the Republic (a highly subjective qualification, which implies that someone could be less compatible with the Republic than the Rassemblement National). In reality, Renaissance will be choosing who will be most likely to vote in favour of future government bills.

Of course, it will work. And we will be entitled to a speech in the same vein as “this vote obliges me”, as quickly uttered as forgotten, and Emmanuel Macron will then consider this vote of opposition to the RN as a vote of support.

But that’s not all. It reinforces, once again, the two-party system that Emmanuel Macron is so keen to see. On the one hand, the democrats, whom he claims as his own, and on the other, a reactionary right-wing party that he has chosen as his adversary. And the left? Let it disappear.

The real objective: to destroy the left

Emmanuel Macron has already taken this extremely risky gamble, and he has always won. He began by absorbing the forces of the right and the right of the left, but that was not enough. A hero is only as good as his nemesis. And that nemesis is the far right.

This narrative construction has enabled it to gradually integrate all the socialists and democrats (including those on the right) into politics, and has seen a major reconfiguration of the landscape on the right, with the Les Républicains party sliding towards the extremes to the point of being hard to distinguish from them, and the emergence of a multi-faceted identity bloc (the Rassemblement National, Reconquête…).

This non-spoken coalition, which Emmanuel Macron regularly reinforces with a host of social disruptions (always a godsend for populists), virilist rhetoric and xenophobic bills, allows him to use democracy to remain at the head of a country that he then rules in a highly undemocratic manner, using 49.3.

Some people wonder whether Macron wants to put the far right into power to prove that it is incapable of governing, while others disagree. For me, it’s not a question of “if”, it’s “when”.

By building up a ‘democratic’ bloc against a ‘reactionary’ bloc from the start of its political history, he is seeking to create an American-style two-party system. With one obvious consequence: at some point, when people are fed up with A, they will vote for B. And that’s what we’re starting to see, with record turnouts.

In the end, Emmanuel Macron’s real enemy, the political current that does not exist in this political vision, is the left, and more particularly the social and ecological left, the only one to oppose capital, which has no enemies on the right (however far to the right you go).

A Nupes hope?

The Nouvelle Union Populaire Écologique et Sociale (Nupes) is a coalition of French left-wing political parties formed in 2022 on the eve of the last legislative elections. It is a union of circumstance, the aim of which is to put in place single candidates in the majority of constituencies. The incumbents will be retained, regardless of their track record, and for the other constituencies, the parties’ results in previous elections will be used to decide who to field.

It’s a necessity, because no matter what you think of this or that left-wing candidate, on the whole, they are democrats. Yes, there are people on the left who discriminate, there are sexual abusers, charlatans, populist influencers and ecociders. But if you know this, it’s because information is circulating. It’s because there are investigations. It’s because there are internal processes and a permeability to journalistic work to enable this transparency, unparalleled on the right and the far right. We don’t see problems because they exist, we see them because the law of silence is not as strong there.

Obviously, we on the left should be cleaning up our act more quickly. Frankly, it hurts me to be a left-wing activist knowing that some politicians, who we know are highly problematic, are up for a new job. Can we wait until we’ve cleaned house before putting the left back in power? I doubt it.

Whether it’s via Emmanuel Macron or the RN, the far right is getting more entrenched every day and the doors are closing from election to election. When we finally have our ideal left in hand, we won’t be allowed to vote for it.

And if the left “wins”?

So let’s be honest for a minute. “Winning”, in this context, will be bettersweet.

Since 2022, in the National Assembly, we have:

  • ~90 RN
  • ~60 LR
  • ~250 Renaissance (presidential party)
  • ~30 LIOT/non-attached members
  • and ~150 Nupes

Knowing that ~290 MPs are needed for an absolute majority, the left is currently not a sufficient force to defend a text. Even if we add Liot and the Non-attached Members.

For the left to really “win” these elections, i.e. to become a real reservoir of votes for the presidential party, many constituencies would have to be convinced that the left is a better alternative than remaining loyal to Renaissance or voting RN. I’d like to see that happen, but in view of what’s just happened in the European elections, I don’t think it’s going to happen.

What is at stake here is the survival of the left, with a RN group that could potentially double in size and constitute the largest pool of votes for passing legislation.

What will the Rassemblement National do?

For the moment, RN MPs have little say in the drafting of the law. On the other hand, their votes are real and should logically be increasingly powerful2.

On wages and living standards, the RN is making the French poorer by voting:

  • against increases in the minimum wage, civil servants’ pay and housing benefit
  • for the tightening of all unemployment conditions (end of rights after giving up a job, tightening and abolition of benefits)
  • against indexation of wages (on the cost of living)
  • against freezing rents or freezing the price of basic necessities
  • against free school meals and supplies
  • against increasing student grants (despite supposedly meritocratic values)

The RN supports the order of capital and high wages, voting:

  • against wage controls
  • against the wealth tax
  • against the abolition of public subsidies for companies that pay dividends without increasing their workforce
  • against the fight against tax fraud
  • in favour of video surveillance

And as a result, the RN votes for anything that weakens and privatises public services:

  • against funding for universities, hospitals, fire brigades, the National Financial Prosecutor’s Office or court clerks ( you betcha…)
  • against lowering VAT on public transport
  • for the abolition of the audiovisual licence fee, leading to a radical overhaul of public broadcasting at a time when it has never been so successful

Finally, the RN is the enemy of social responsibility and ecological transition, voting:

  • against the conditioning of public aid to companies complying with ecological constraints
  • against regulating the installation of doctors in areas where there are already enough of them
  • against gender equality measures such as increasing women’s access to positions of responsibility in the civil service
  • against the vegetarian option in canteens (even though it would reduce costs for all children)
  • against abolishing the niche tax on aviation fuel
  • against a climate wealth tax
  • against taxing yachts and private jets
  • against the obligation for companies financed by the France 2030 plan to publish a carbon balance sheet
  • for the exclusion of kitchen utensils from the ban on eternal pollutants (PFAS)
  • against combating sectarian organisations
  • for oil-fired heating, for limiting the use of wind turbines, including in the open sea, and against the compulsory installation of solar panels on large car parks;
  • and against all forms of inclusive writing.

Finally, during the debate on the pension reform, their opposition was superficial. They applauded the conclusion and even though in their programme they propose retirement at 62, it is under the guise of drastic conditions. When it came to the vote on the revaluation of small pensions, they voted… against.

Regularly asked what they would do with the reform, they announce that they would go back on it… well, if France’s debt level allows it (hint: once in power, they’ll say they won’t).

Paradoxically, the RN is not proposing anything to improve the level of education, which is directly correlated with income and therefore future pensions. Marine Le Pen has repeatedly insisted that young people should enter the market more quickly, which would automatically reduce the costs of public higher education. It’s a constant for the far right all over the world: they don’t fund institutions likely to train opponents.

So what’s in store for us isn’t exactly glorious: more inequality, dismantled public services, a debt that keeps growing because of a lack of investment in long-term projects like the ecological transition, and at the next elections, once again, the choice between the plague and the cholera.

There’s only one solution: vote for the left, now and sustainably.

  1. In France, the Rassemblement National (far-right) came out on top with 31.37% (30 seats), followed by the presidential party (liberals) with 14.60% (13 seats), the socialists (social-democrats) with 13.83% (also 13 seats) and then the Insoumis (lef) with 9.89% (9 seats). ↩

  2. All data are available on datan.fr which itself links to the National Assembly website. ↩

Future Of Tech
the challenges of digital tech… without it

At the end of last week, I spent 2 hours with Gaëlle Vandenberghe’s class at the Philippe Cousteau High School1. I co-facilitated a workshop with about twenty teens, organised around a card game that summarised the major challenges of digital technology through the construction of a mobile application.

Future Of Tech is a workshop proposed by the association Latitudes, supported by the French Ministry of Education, Réseau Canopé and several education authorities. Latitudes takes care of the logistics as well as the training of the facilitators, following a very well-constructed curriculum that I followed throughout February, comprising a dozen themed modules and challenges to be taken up within the community.

A colourful card bearing the words in French "Future of Tech: the workshop that introduces high school students to the professions and challenges of digital technology".

The cards are really fun.

The Future Of Tech workshop deliberately emphasises the use of non-digital media (cards) to encourage reflection, and to emphasise the role of digital technology in social issues rather than technological innovation. The workshop looks at tech careers, but above all seeks to awaken the critical minds of teenagers about its social and environmental components.

During the workshop, the teenagers learn about the digital professions, but they also talk about commitment, diversity and designing technologies that are more respectful of people and their lifestyles, accessible to all, sustainable and encouraging citizenship.

I loved taking part in this first workshop and knowing that I had the support of my employer, Contentsquare, which every year offers its employees the chance to carry out voluntary work. We recently adopted the Benevity platform and I was delighted to discover, after starting my training as a facilitator, that Latitudes was listed on it.

I’ve contacted several local high schools to repeat the experience and see what other questions and ideas emerge from young people who want to and can change the world. I also signed up for training courses for other Latitudes workshops, designed to run in unis and entreprises (Bataille de la Tech, Bataille de l’AI). This is so exciting!

  1. There’s even an article on the school blog (in French). ↩

En retrait

Gender equality. We’re not there yet. Yet it is an ethical1, social2 and economical3 imperative. But to get there, we need to decentralise the world from men.

Current political approaches all have the same flaw: they look at gender equality through the prism of men. Even when it comes to measuring inequality, we continue to measure it using men as a reference point.

Imagine you’re a man walking along a winding road and a female friend who also knows the way is following you. You want this friend to arrive at the same time as you, so you turn around often to check that you can see her, and to gauge the distance between you. There are several ways of looking at this situation.

  1. Either you choose to stay ahead and spend your time assessing the distance, then constantly re-accelerate at the right moment to stay ahead.
  2. Or you decide to stop for a few seconds while your friend joins you and you move forward hand in hand, but this means sacrificing part of your lead.

Except that sometimes the path isn’t just winding. It’s also narrow, and only one person gets through. To achieve equality between women and men, men have to agree to let women go in front of them occasionally and temporarily. And too bad if they speed up afterwards and overtake us, we’ll deal with that after.

Otherwise, even with the best will in the world, we men will always have women in our rear-view mirror.

  1. Human equality is a fundamental principle of justice. Every individual must have the same rights and opportunities. It is a matter of respect and dignity. ↩

  2. The emancipation of women is important for their personal fulfilment and the fight against all the consequences of discrimination, such as sexist and sexual violence, psychological violence, educational violence, but also - and we easily forget this - economic violence (control of money, access to accounts) and destruction of property. ↩

  3. Equality between women and men increases diversity in companies and enriches creativity, improves work organisation and enhances the financial prospects of non-heteronormative households. ↩

AI and LLMs, uses I see in 2024

When the hype started to build around what is commonly referred to as “Artificial Intelligence” (AI), I wrote a first article to highlight some of the opportunities and risks I saw in it. The market has matured, and so has our perception of the world. What more can I say in 2024 from what I can see (a very limited vision, I’m aware).

In terms of communication, the general and specialized press continues to wallow in confusion by naming everything and anything “AI”, which generates an invitation to companies to jump on the bandwagon and, in turn, announce that they are doing AI without it really being clear whether they are using algorithms, machine learning or deep learning.

The public understands “software that makes decisions without us being able to question its inner workings”. I suppose that once a certain lack of transparency is reached, it doesn’t matter how the sausage is made.

Any sufficiently advanced technology is indistinguishable from magic. Arthur C. Clarke

Image description, we’re there. Cloudinary automatically generates captions for images for which descriptions have not been provided at upload (demo). Ice Cubes, Mastodon’s iOS client (in the App Store), offers on-the-fly image caption at upload time, to help with entering alternatives. The official Mastodon Web app does it too.

Translation, we’re there too. With all the unfortunate consequences that entails: the giant Duolingo, for example, has massively shed its contract staff. No need for translators when you can rely more heavily on a software layer providing an output that is probably less qualitative, but so much cheaper.

All professions involved in the production of graphic and textual content for the general public are undergoing profound changes, and therefore are in crisis. Workers are, even more than before, competing on their ability to develop the most precise prompts to generate relevant content, carrying a recognizable message and style. Freelance or commissioned production is disappearing, replaced by automated generation just good enough to create advertising opportunities.

On the B2B and government side, tools for reformulating and summarizing files are beginning to appear. Without masking the complexity of a file, and by providing references to the elements it summarizes, hence enabling verification, these tools could rapidly improve the performance of all economic agents: real estate, property management, accounts management, consulting, and so on.

On e-commerce marketplaces and content sites for netlinking, thousands of external references are integrated and reformulated on the fly, ensuring that search engines don’t label the content as duplicated. And search engines themselves are getting in on the act, reserving their indexing capacity for sites that accept that their content be stolen and never displayed, since it will have been reformulated by a conversational program that, if it’s polite enough, will add a link as a footer reference.

After years of believing in the Net Promoter Score when every serious study confirms its invalidity1, web operational teams are gradually moving away from the concept and returning to verbatim analysis, but above all to behavioral analysis. User frustration doesn’t necessarily translate into feedback (verbatim or note), but it can be collected and correlated with friction in the achievement of objectives.

A number of design aids are emerging within prototyping or A/B testing solutions. Based on an automatically identified of a deviation from the norm, they try to suggest ways of optimizing the UI. I have my doubts about this approach. At best, guided by an obfuscated internal rule engine, this Computer-Aided Design (CAD) would influence the neutralization of the brand image, by proposing highly consensual UIs that are not necessarily more usable or accessible. In the worst case scenario, proposals resulting from automated learning without business context will be in total contradiction with the design system and the organization’s objectives.

The situation is a little different for the training, advanced business support and technical documentation professions. Having always had a niche clientele, they are not interested in producing huge volumes of answers, but in accurate content. Reducing LLM training to a very small documentary base, they can quickly benefit from conversation aids. As a user of a support feature, you’re not necessarily talking to a conversational bot, but the person handling your need is sure to have a writing aid that enables them to gain in productivity, while validating answers. Some highly specialized conversational bots may even provide precise answers to niche thematic questions.

Once properly trained, these assistants can operate with little supervision. Thus, 2024 should see the return of shopping assistants triggered on demand or after a frustration signal is given. By re-engaging users, they can only save a few sales.

The qualification of these signals is a major emerging theme in e-commerce. Supported by data collection that is necessarily consented to (at least in Europe), it is now the basis for a personalization of experiences, made technically possible by the composability of the e-commerce experience. This evolution should explode the well-established UI norms (the home page, the product list page, the product page) to allow for more audacious compositions. We’re dealing here with an in-depth evolution rather than a passing fad, which began years ago with the static movement and headless e-commerce and continues today with custom composable architectures (MACH) or solutions from Adobe or Salesforce that are constantly pulling the market towards personalization.

Of course, in this galaxy of “signals”, there are both wheat and chaff. I’m very skeptical about differentiating users on the basis of behavioral profiles determined by analyzing a few actions at the start of a session. I’m not saying it’s not possible, but I’d very much like to read the scientific research attesting to the correlations between the models constructed and another form of perception of reality. It sells very well, web operators really want it to be true, but I’ve never seen the data to back it up.

In any case, the market is probably getting too carried away with personalization. Before being fully automated with dynamic signals, dimensions will certainly be chosen manually, probably by merchandising teams. Among emerging uses, I foresee the acceleration of product image personalization during campaigns. I expect to quickly find sites where my “For You” selection will contain products whose image will have a predominantly pink palette. Once again, Cloudinary dominates the market with a proposal for on-the-fly modification of packshots from an instruction, but other Digital Assets Management companies should quickly follow suit if the proof of value is consistent with the cost. Indeed, the production of personalized images remains an energy-intensive task, and who can say if the small number of retailers able to afford these tools will make them profitable.

The last area I see emerging, driven by “AIs”: in-demand personalization. Today, natural language analysis tools make it possible to understand a sentence like:

I want to consume products made in France and produced in an eco-responsible way.

and not only filter the catalog accordingly, but also generate part of the product description in an adapted manner. For example:

Manufactured at Choulat-Le-Tréfon, in the heart of the French Champignois region, from recycled tire inner tubes, this belt is delivered according to the EcoStandard®.

A good way for huge marketplaces in need of an image to recreate a link with their customers, while at the same time showing a form of respect for their specific expectations.

  1. Amoung other issues, NPS is based on the principle that a person cannot be both a promoter and a detractor. But when someone takes the time to leave you a rating and a negative review when they have the choice not to, it’s clearly because they like your brand. Your NPS is only ever a reflection of the people who are not indifferent to your brand. Other people are not taken into account. Learn more: “Where Net Promoter Score Goes Wrong”, Christina Stahlkopt. ↩

In protest of the pension reform

I will be on strike and demonstrating tomorrow, to press on the French governement for the withdrawaof the pension reform bill.

Today, we need to rethink the respective places of work and free time in our lives, giving us back time for ourselves and for others.

I dream of one day being a grandfather who is available to look after his grandchildren, go on vacation with them and help my kids with the challenges of their adult lives. And I want to know that there is another way to do this than to become a rentier.

I owe it to my children to defend their legacy. So I hope we’ll get there.

Legacies

As I write, France is on strike at the call of an inter-union coalition (which rarely happens) to demonstrate against the government’s bill on pension reform.

Contrary to what the members of the presidential majority say, this reform is not a necessity to maintain our pension system or its level of compensation. It is not a project related to the increase in life expectancy or that in good health. It is a project of financiers, intended to continue the ongoing reduction of corporate taxation.

And I’m not the one saying it, it’s in the 2023 finance law, forced by a 49-3 in the Parliement (49.3 is a mechanism that allows the French prime minister, following discussions with cabinet (Conseil des ministres) to unilaterally pass any bill concerning financial or social security issues):

Social security administrations will participate in controlling expenditure, notably through pension reform, reform of unemployment insurance to promote full employment, and control of health care expenditure (the ONDAM will increase by 2.7% over 2024-2025 and 2.6% over 2026-2027). Local administrations will also be involved in this control of expenditure, with a target of reducing their operating expenditure by -0.5% in volume each year.
[…]
Thus, the continued deployment of the “France 2030” plan will stimulate the economy and support potential growth by accelerating the ecological transition, promoting investment, innovation, social and territorial cohesion and ensuring digital and industrial sovereignty.
[…]
This containment of expenses will make it possible, without compromising the objective of normalizing the public accounts, to continue the strategy of reducing taxes and social security contributions initiated under the previous mandate in order to support the purchasing power of the French and the competitiveness of French companies.
Projet de loi de finances n°273 pour 2023 , translated in English

France as a start-up: a country that exists only as much as it appeals to the eyes of investors and the wealthy, and that will do everything to ensure that their wealth, which is growing at an ever increasing rate, continues to absorb the trickling riches of our country.

So they repeat the old stories. They lie about the real state of the pension system, they remind us of the public debt “that we will have to pay back” (that’s not that simple), they accuse the strikers of wanting to paralyze the country… the usual tired tales.

Everything, as long as no one ever reminds us that all those things that are being conceptually and then concretely precariousized to the point of being privatized: education, health, logistics, public transportation, culture, security (including income when unemployed, disabled or retired)… are also legacies.

Legacies that obviously we will not pass on to our children.

Volunteer Day

On Friday, I’ll teach computers. Not computer science, just how to do basic stuff with files and a spreadsheet.

To you, it may be nothing, but to the staff and residents of the non-profit retirement home I’m helping, it’s huge. And Contentsquare’s a part of it.

I benefit from the Volunteer Day initiative announced Tuesday by Kat Borlongan, Contentsquare’s Chief Impact Officer. Each company employee can take one day out of the year to support an organization whose primary goals are #philanthropy and social welfare.

As part of Volunteer Day, some of my colleagues take time to collect funds or items for Ukraine or to welcome refugees into their homes. Others tutor or counsel teens in crisis, clean up local parks or participate in street outreach.

Most of us won’t do this for just one day a year. But knowing that the company supports us and decides to maintain our pay for one of these days is a very strong sign of responsibility and commitment to the #community, but also to each member of the CSquad.

That’s not all. Contentsquare has also bootstrapped a Gift Matching Program to amplify CSquad’s donations. I decided to donate €200 to UNHCR, the UN Refugee Agency. Contentsquare matched that donation 1:1, giving another €200.

I’m so grateful and I just wanted to say it. 🙏

Trickle-Down

Some economists use the concept of trickling down to justify actions that allow wealthy people to increase their wealth. The idea behind this is that by investing their fortune, the wealthy would share it with others, who in turn would consume goods and services that would allow them to share their wealth with less fortunate people, etc. This is obviously not true, but I would like to draw your attention to the name of the concept, “trickle-down”.

Trickle-down, in hydrology, is called “surface runoff”. It describes a phenomenon in which water flows from places where it is scarce (mountains, land) to areas or bodies of water where it exists in large volumes (lakes, rivers and, of course, oceans). As the water flows, it carries along particles of varying size depending on the amount of water in motion and the slope. This erosion depletes the soil of its essential nutrients, making it less fertile. If the runoff continues for a long period of time, then particles of soil or stone are actually washed away, which can lead to long-lasting gullies on the land, which can further dry out the soil.

This is an excellent description of capitalism.

Capitalism describes a phenomenon in which capital flows from where it is scarce (the poor), to funds or accounts where it exists in large volumes. Each time capital flees to the wealthy, it takes with it essential elements of financing the commons (culture, health, transport, education, unemployment, retirement…). This erosion depletes society of its essential nutrients, making it less fertile. If the stream continues for a long time, then public services start to disappear, which can lead to lasting degradation in society, which can continue its impoverishment for the benefit of capital.

My Privileges

My privileges are numerous and allow me to appreciate my life, to better understand what the lives of others may be like, and to put into perspective the few trials that I have encountered so far and the merit that I may think I have in overcoming them. The exercise of listing my privileges is delicate and imperfect. By doing so, I will forget things. I do not forbid myself from revising this list later.

Education and Training:

  • I have pursued studies that corresponded to my desires.
  • I have completed higher education that gives me access to desirable jobs.
  • The costs related to my education have been covered by my parents and by the community, and from a young age, I have had access to extracurricular activities and all the support I needed.
  • I can have a serious conversation with any English speaker.
  • I work in a field that I enjoy.

Health and Well-being:

  • The costs related to my health have been covered by my parents and by the community. I have access to a doctor and to the medications necessary for my health and that of my family.
  • I have never been in depression. I have never been mentally incapacitated. I have never wanted to commit suicide.
  • I have never had an addiction.
  • I am always comfortable with my gender, which is my birth gender, and this has never been questioned by those around me.
  • Society has never closed doors on me because of my physical abilities. I live without suffering from the negative gaze of society on my body.
  • I have always been monitored and cared for, without fear of not being able to pay my out-of-pocket costs.
  • I have always had access to healthy and varied food.

Safety and Freedom:

  • No one has ever told me that I could not do something for reasons independent of my health or over which I had no control (my origins, my sex, my gender…).
  • No one has even told me that it would be more difficult for me than for others, that I should be more careful than others for my own safety.
  • I have never been attacked or aggressed because of my lack of religious beliefs.
  • My identity has been checked 4 times since I became an adult.
  • I walk in the street without fear or being bothered.
  • I have almost never been reduced to a stereotype or a class and judged negatively solely based on this belonging.
  • I have never been afraid that people would know my sexual preferences: neither my parents, nor people in the street. I am comfortable with my heterosexuality and with my gender expression, which seem logical and natural because their articulation corresponds to the norm, which allows me to be comfortable with public displays of affection with my wife or my children, without fear for her, them, or me.
  • I am rarely reduced or brought back to my appearance. I have the leisure of not being conscious of my appearance for a large part of my day.
  • I have never had to hide who I am or what I believe in. My survival or even my physical integrity has rarely been endangered in public space.
  • I am free to express my political opinions without consequences to my health or safety. The impact of this expression on my professional environment is not zero, but it is not critical either.
  • I am free to engage in the causes that are close to my heart, without fear of reprisals (but sometimes a little of repression)

Work and Finances:

  • My profession is valued by society; it is well paid. Its arrangements are practical and allow for teleworking. Demand is strong; finding work is easier than in other fields.
  • I have a permanent job, with a fixed gross salary that I can count on.
  • As a man, I am statistically better paid than my female colleagues with equivalent skills, and I have more facilities to obtain better-paying jobs.
  • I have always lived in a house with access to hot water. I have rarely been cold, and never because I could not pay for heating. I have never been afraid of not being able to pay the rent.
  • I have benefited from social services to house me or to welcome my children at times when I did not have the means or skills to do so.
  • I live in a country that maintains laws allowing me to regularly take paid vacations, to have free time, and to rest.

Relationships and Social Position:

  • When I speak, people listen to me.
  • When I speak, people believe me.
  • I am not alone. People are there for me and have the means to help me if something serious happens to me, whether it is my family, my friends, or professional acquaintances.
  • My parents, as well as those of my wife, are still alive. They support us and are present for our children.
  • I was 6 years old the last time someone made comments about my accent.
  • I am white, and in the environment in which I evolve, it is considered the norm. I am not noticed; I am not ostracized for my skin color. I am never the only one “like me” in a space.
  • I have never felt declassed or useless.
  • I live without suffering from the negative gaze of society on my body.

Travel and Culture:

  • I have traveled to several countries and met several cultures. I have never been hindered in my movements because of who I am or where I come from.
  • I was born in a country with a higher standard of living than others, which makes travel easier. I can spend 3 weeks of vacation in Cambodia; a Cambodian cannot easily spend 3 weeks of vacation in France.
  • I have free and sometimes even free access to knowledge, culture, and sports facilities, whether it is libraries, shows financed or subsidized by the community. I have the means to obtain what the community does not provide.
  • I have often lived in the city, which has allowed me to access many services with public transportation. I have also benefited from parks and green spaces maintained by the community.

'How to apologise properly', translated

This tutorial by Mélanie Wanga on how to apologise is a gem. I’ve listened to it at least 200 times. It’s short, to the point and it’s helped me, more than anyone else, to question the way in which I used to apologise, both in the professional world and to those close to me.

I highly recommend it and I decided to provide an English translation (below).


For all those who know they’ve screwed up but have never learned to apologise, or who aren’t sure how to do it, here are a few recommendations.

What not to do: minimize your actions.

To the people who have felt targeted here or elsewhere for the past eleven years due to one or more of my mocking remarks…

Boom 🎶

When you’ve hurt someone, you’ve caused pain. Your action has led to suffering. So now isn’t really the time to downplay that act with linguistic twists that distance you from reality.

In the case of the “Ligue du Lol,” it involves harassment, denigration, and persistence. Therefore, these are the terms the “loleurs” should have used in their apologies, rather than expressions like “awkward remarks” “targeted obsession” or “mocking jabs” which are completely inappropriate.

What to do: acknowledge what you’ve done

Yeah 🎶

The person you’re apologizing to expects you to name the facts. Be straightforward about them. When you step on someone’s foot, it’s not accidentally pressing on the end of a person unintentionally. You stepped on someone’s foot.

The injured person needs to see that you understand what happened, what took place. So muster your courage, even if it scares you, and name what you did.

I harassed you.
I attacked you.
I hurt you.
I lied to you.

What not to do: taking responsibility away

Back then I was young. Twitter was something else. I have the impression that I was a hater, a loner rather than a leader. I felt I belonged to something.

Boom 🎶

Taking no responsibility was the main flaw in the apologies made by the members of the Ligue du Lol. They’ve covered everything: their age, their geographical origin, their frustrations… but this has nothing to do with their participation in mass harassment. Thousands of people are in these situations and yet they didn’t harass anyone.

Taking responsibility away is when you feel too guilty about what you’ve done and you start looking for reasons why you acted as you did. And… that’s not what an apology is about. The apology is about the other person, not you.

What to do: recognise your responsibility

Yeah 🎶

If you’ve done what you’ve done, it’s because somewhere, deep down, you wanted to. Whatever the environment, you are the person responsible.

You have to accept this and not look for extenuating circumstances because they are not what the person you are apologising to expects, nor what they need.

It’s all my fault, I took that decision and I regret it.

What not to do: talk about your personal context.

At the time, I came from the suburbs and had no contacts in Paris. I already loved stand-up and, frustrated at not being able to do it, I was pouring out my jokes on the internet.

Boom 🎶

Your personal circumstances aren’t the point, in fact, certainly not when you spend more time detailing them than making a genuine apology.

What to do: talk about the effects on the victim

Yeah 🎶

A good apology speaks of the destructive effects of the act on the victim. Think about it and ask yourself what your action took away from the person: their well-being, their confidence, their joy… What were the effects?

If you’re struggling, now’s the time to empathise and imagine how you would feel if you were in your victim’s shoes.

I understand that what I did prevented you from doing this or that. I imagine you felt humiliated, betrayed, isolated.

What not to do: clear your name

To anyone who may have felt affected.
The way you experienced things.
To those who are still suffering from trauma.
To anyone who might have been affected.

Boom 🎶

First thing: don’t say you are sorry. You’re apologising to your victim, it’s not about yourself.

And another classic mistake is that you don’t apologise if you’ve hurt someone. You apologise for your actions. Your apology is sincere and has value because you are questioning your act in itself and you are sorry for having committed it, whatever the consequences. So you apologise unconditionally.

On the other hand, you don’t blame others for the consequences, as if the problem were the victim’s perception, reactions or feelings.

What to do: offer a genuine apology

I ask for your forgiveness.
I apologise.
Please forgive me.
Please excuse me.

Yeah 🎶

All these expressions highlight what is happening. The fact that you have realised that you have committed a hurtful act, that you have understood its impact and that you are apologising.

On the other hand, your victim’s forgiveness is not owed to you.

All you can do is offer your sincere apologies as best you can. And then you can promise not to do it again.

Whose territory am I?

A map is not the territory it represents, but, if correct, it has a similar structure to the territory, which accounts for its usefulness. Alfred Korzybski, Science and Sanity, p. 58.

I know I’m tilting at windmills but I also think this needs to be reiterated.

We know that an increasing number of private and public organizations are modeling us. Online tracking, connected devices, ghost profiles, cohort analyses… We are incompletely defined, wrongly (in both senses of the word) mapped.

What we still have to understand and assimilate in Popular Culture is the usefulness and use of these maps by and for those who design them. Because in the same way that a path indicated on a map becomes more frequented, and therefore more clearly marked on the territory than a non-mapped path, the aberrations of our models become, due to the manipulations of our publication timelines and filtered bubbles, our own mental pathways.