When Foundations Tremble: Data, Algorithms, and Old Certainties

When Foundations Tremble: Data, Algorithms, and Old Certainties

A day when the Swiss Confederation discovers its leaks, when AI promises to replace mathematicians and developers, and when everyone wonders what is still standing. Between ransoms, quotes and curfews, reality reminds the well-meaning of its presence.

Let us begin with the leaks that cause anger: Thierry Weber relays the case of the Akira gang, which is demanding a ransom for the data of more than 100,000 employees and former employees of the Confederation, which leaked at PK Softech, a software supplier for pension funds. Here is a chain of trust that breaks at a discreet link, and one suddenly realizes that a state’s security depends on providers whose names nobody knows. On a completely different note, Caroline Mignaux observes that consumer society drives people mad, while Isabelle Bouilloux raises the question of whether artificial intelligence can replace mathematicians, drawing on Terence Tao’s analysis. Three ways of saying that our systems, economic, cognitive or digital, are being called upon to prove that they still hold.

Trust, precisely, is also expressed in quotes and perceptions. Laurent Babikian invokes Nelson Mandela and his idea that hatred and love are learned, a way of recalling that none of this is a given. Pascal Wagner-Egger points to a fascinating article by his colleagues and friends Peter Barrett and Antoine Marie on violence in France and the perception of protests, which resonates with current events. And Hugo Clément opens the debate: to defend women’s cause, should pornography be banned? Three entries, one and the same question about what we think we see and what we want to ban, between teaching feelings and the politics of the gaze.

The gaze is precisely the journalist’s trade, and Amèle Debey recalls that their role in relation to the authorities raises fundamental questions of transparency and counter-power, especially when crucial information is presented without context or figures. Lionel Girard, for his part, wonders whether ChatGPT is preparing a world without Amazon, which shifts the tug-of-war to the side of commerce. Franck Rivat extends the reflection in a darker tone: is this urban and economic war in France set to last on weekends until Christmas, or until the next presidential election, in the wake of Moody’s negative rating? Three ways of asking who controls the narrative, from the television set to the shopping basket.

Machines, for their part, do not ask anyone’s permission. Wilfried de Renty warns that banks are beginning to lose control as AI agents become established. Pierre Jean Duvivier dit Sage ironically comments on the “great cause” of women’s rights under Macronie, listing pell-mell the Bruel affairs, the first lady, a justice system that convicts only a tiny fraction of rapists, and toxic products launched at protesters. And he mocks CNEWS, which comments in Paris on what is happening in Tirana, pointing out that staffing levels apparently are not enough to guarantee lucidity. Technology advances, politics stumbles, and the sports commentary of TV panels becomes a combat sport.

Because the web has its mysteries, and Sage tells a fine one: on October 21, like many IT professionals, he was trying to patch his PHP servers after the discovery of new vulnerabilities, before finding that his Debian machines didn’t have the right repositories, which brings him back, as always, to Ondřej Surý, that name that has recurred for twenty-five years in his life as a developer without his ever having seen his face. He draws from it a reflection on an industry that turns over hundreds of billions and rests on a few enthusiasts, sometimes volunteers, from whom everyone benefits and whom no one knows, with this dizzying question: what would happen if these people stopped one fine morning? Sean B., for his part, published on GitHub a weaponized exploit for the two CVEs that landed that day on PHP, targeting the latest WordPress, with an “Enjoy!” that needs no comment. The hand that repairs and the one that breaks sometimes cross paths on the same repository.

AI is no longer content to code patches: Camille Roux reports that Epic Games is going “AI first” on its Marvel/Star Wars shooter with Disney, no more code written by hand, models like Claude generating the code and guiding the assets. General Jean Pierre Meyer, for his part, discusses intelligence, command and decision-making in the age of AI, proof that the subject is leaving the studios for the general staffs. And Sage, in a more prosaic register, is annoyed by Gabriel Attal, whose merit he disputes in view of a diploma “adjusted” at Sciences Po and virtually no professional experience, recalling that intrinsic incompetence has consequences for far more people, and with our taxes. From algorithms to résumés, the question of competence remains the same.

What remains is the body, that great forgotten element of org charts. Two Silicon Valley billionaires are raising 200 million to create genetically modified war dogs, reports Futura, which gives an idea of what humanity finds to finance when it is bored. Vincent Flibustier, for his part, recalls that when you run a blurred image through ChatGPT, the model literally invents the pixels and fabricates vaguely resembling faces, potentially compromising. Philippe Douale, finally, shows a visual where the post displayed on screen belongs to someone else, and where the comments become a boomerang of visibility. Blur, decidedly, no longer protects anyone.

There are also those who look at crowds with the weary eye of a biologist. Sage notes that working before 10 a.m. would go against our biological rhythm, citing researcher Paul Kelley of the University of Oxford, for whom traditional schedules disconnected from our internal clock cause chronic sleep deprivation, with fatigue, concentration problems, anxiety, weakened immunity and cardiovascular risks, and a study published in Frontiers in Human Neuroscience showing that a 10 a.m. start to classes improves the health and results of students aged 13 to 16. The same author observes that the MAGA phenomenon illustrates a crisis of Western literacy, where ignorance would become a virtue. And he summons the memory of a film about management, human resources and corporate culture to describe a youth that refuses to become a cog, while recalling that the world, absurd and brutal, remains less terrible than it once was, and that the educational failure perhaps consists in having stopped learning to confront complexity rather than to protect oneself from it.

On the political front, the cup overflows. Sage argues that complacency has a cost the country can no longer bear: permanent curfews, injured police officers, severed hands, weekends lost for shopkeepers, tourists fleeing, businesses at a standstill, and, according to him, a consensus from LFI to the RN to tell Macron to leave. He adds that there is nothing left to save in this regime, which is a way of closing the chapter rather than writing it.

The protest, for its part, does not stop at borders. Sage observes that after France, Spain, Portugal, Italy, Albania, India, Argentina, Serbia is now ablaze in turn, and wonders whether our societies produce anxiety-inducing effects common to countries that are nevertheless very different, with no obvious political link between socialist Spain, Meloni’s Italy and Serbia. He also returns to CNEWS commenting on three minutes of footage filmed on the left in Paris and on the right in Albania, Albanian flags included, to draw from it a demand for a ban or a sanction. Alexandre Poidatz, for his part, praises apt words and a powerful turn of phrase regarding Villepin, Le Pen and deportation, proof that language remains a battleground.

There remains a question of a visa and a green card for Indians, which Mario Cotza raises bluntly, a way of recalling that the great migration concerns also play out in forms. After the ransoms, the exploits, the quotations and the curfews, the day ends on an administrative form: that is perhaps the true face of our societies, less spectacular than their crises, but just as decisive.

My bestiary of the day

The voices and pens crossed in this edition:

Editorial summary produced with the assistance of DeepSeek from the publications listed above.

The post When foundations tremble: data, algorithms and old certainties appeared first on the pens that bite.

Frequently Asked Questions
FAQs When Foundations Tremble Data Algorithms and Old Certainties

1 What is When Foundations Tremble about
Its about how data and algorithms are shaking up longheld beliefs and established ways of doing thingsand how we can respond when the certainties we relied on stop holding up

2 What does old certainties mean here
The assumptions rules of thumb and expert judgments people trusted for yearsthings like experience knows best or this is just how its donethat now get challenged by datadriven evidence

3 Why are the foundations trembling now
Because cheap computing huge datasets and powerful algorithms let us test and question things at a scale and speed that wasnt possible before

4 Whats the difference between data and algorithms
Data is the raw informationnumbers text records An algorithm is a set of steps or rules that processes that data to produce a result like a prediction or a ranking

5 Do I need to be technical to understand this topic
No The core ideaswhat data shows what algorithms do and why old assumptions break downcan be understood without coding or math

6 What are the benefits of relying more on data and algorithms
Better decisions at scale spotting patterns humans miss consistency speed and the ability to test beliefs instead of just asserting them

7 What are the most common problems
Bad or biased data algorithms that are right on average but wrong for specific groups overtrusting models and losing human judgment in the process

8 What is garbage in garbage out
It means the quality of an algorithms output depends on the quality of its input data Flawed data produces flawed results no matter how sophisticated the model

9 What is algorithmic bias
When an algorithm systematically produces unfair or skewed resultsoften because the training data reflected past biases or doesnt represent everyone

10 Can you give a realworld example of old certainties being overturned
Yesfields from medicine to hiring have seen longtrusted methods challenged by evidence

11 Does this mean experts and experience no longer matter
No Data and algorithms

Related Posts