Vue lecture

Google: Don't Make 'Bite-Sized' Content For LLMs If You Care About Search Rank

An anonymous reader quotes a report from Ars Technica: Search engine optimization, or SEO, is a big business. While some SEO practices are useful, much of the day-to-day SEO wisdom you see online amounts to superstition. An increasingly popular approach geared toward LLMs called "content chunking" may fall into that category. In the latest installment of Google's Search Off the Record podcast, John Mueller and Danny Sullivan say that breaking content down into bite-sized chunks for LLMs like Gemini is a bad idea. You've probably seen websites engaging in content chunking and scratched your head, and for good reason -- this content isn't made for you. The idea is that if you split information into smaller paragraphs and sections, it is more likely to be ingested and cited by gen AI bots like Gemini. So you end up with short paragraphs, sometimes with just one or two sentences, and lots of subheads formatted like questions one might ask a chatbot. According to Google's Danny Sullivan, this is a misconception, and Google doesn't use such signals to improve ranking. "One of the things I keep seeing over and over in some of the advice and guidance and people are trying to figure out what do we do with the LLMs or whatever, is that turn your content into bite-sized chunks, because LLMs like things that are really bite size, right?" said Sullivan. "So... we don't want you to do that." The conversation, which begins around the podcast's 18-minute mark, goes on to illustrate the folly of jumping on the latest SEO trend. Sullivan notes that he has consulted engineers at Google before making this proclamation. Apparently, the best way to rank on Google continues to be creating content for humans rather than machines. That ensures long-term search exposure, because the behavior of human beings -- what they choose to click on -- is an important signal for Google.

Read more of this story at Slashdot.

  •  

Viral Reddit Post About Food Delivery Apps Was an AI Scam

A viral Reddit "whistleblower" post accusing a major food delivery app of systemic exploitation is "most likely AI-generated," reports the Verge. From the report: The original post by user Trowaway_whistleblow alleged that an unnamed food delivery company regularly delays customer orders, calls couriers "human assets," and exploits their "desperation" for cash, among other indefensible actions. Nearly 90,000 upvotes and four days later, it's become increasingly clear that the post's text is probably AI-generated. Considering the delivery app industry track record of exploitation of its drivers, it's easy to see why so many people believed this was the real thing. The Verge put the original 586-word Reddit post through several free online AI detectors, in addition to Gemini, ChatGPT, and Claude. The results were mixed: Copyleaks, GPTZero, Pangram, Gemini, and Claude all pegged it as likely AI-generated, but ZeroGPT and QuillBot both reported it as human-written. ChatGPT played it down the middle. Reached by The Verge on Signal, Trowaway_whistleblow provided an image of an Uber Eats employee badge. That image was generated or edited with Google AI, according to Gemini. The image shows an Uber Eats logo above two black boxes, presumably covering an employee name and photo, and the words "senior software engineer." It's odd that an engineer's badge would have the Uber Eats logo, and not the Uber logo, according to Gemini. That, in addition to slightly misaligned words and warped coloration at the edge of the green border, are reasons Gemini thinks it's inauthentic. (Uber later confirmed that Uber Eats-branded employee badges do not exist.) "Not only are the claims fake, but they're also dead wrong," Uber spokesperson Noah Edwardsen told The Verge. Uber Eats' Andrew Macdonald wrote on X, "This post is definitively not about us. I suspect it is completely made up. Don't trust everything you read on the internet." DoorDash CEO Tony Xu also denied the redditor's "appalling" allegations. "This is not DoorDash, and I would fire anyone who promoted or tolerated the kind of culture described in this Reddit post," Xu said in a post on X.

Read more of this story at Slashdot.

  •  

Capture de Maduro : SpaceX active Starlink d’urgence au Venezuela dans des circonstances floues

starlink maduro venezuela

L'opération militaire des États-Unis au Venezuela, qui a mené à la capture de Nicolás Maduro le 3 janvier 2026, a aussi perturbé les accès à Internet à Caracas. En réponse, SpaceX a annoncé l'activation de Starlink. Une manœuvre inédite, car le service n'a théoriquement aucune existence légale dans le pays.

  •  

Finland Seizes Ship Suspected of Severing Undersea Cable To Estonia

Finnish authorities on Wednesday seized a vessel suspected of severing an undersea telecommunications cable that connects Helsinki to Tallinn by dragging its anchor across the Gulf of Finland, the latest in a string of infrastructure incidents that have put Baltic Sea nations on edge since Russia's 2022 invasion of Ukraine. Police are investigating the case as aggravated criminal damage and have not disclosed the ship's name, nationality or details about its crew. The cable belongs to Finnish telecoms group Elisa. Estonia's justice ministry reported that a second telecoms cable connecting the two countries -- owned by Sweden's Arelion -- also went down on Wednesday. This follows Finland's December 2024 boarding of the Russian-linked oil tanker Eagle S, which investigators said damaged a power cable and multiple telecoms links using the same anchor-dragging method. A Finnish court in October dismissed criminal charges against the Eagle S crew after prosecutors failed to prove intent.

Read more of this story at Slashdot.

  •  

Revue de presse de l’April pour la semaine 50 de l’année 2025

Cette revue de presse sur Internet fait partie du travail de veille mené par l’April dans le cadre de son action de défense et de promotion du logiciel libre. Les positions exposées dans les articles sont celles de leurs auteurs et ne rejoignent pas forcément celles de l’April.

[Goodtech] «Libre à vous!»: découvrez quel prix spécial du jury a reçu cette émission de radio

Le jeudi 11 décembre 2025.

L’émission de l’April, «Libre à vous!», vient de remporter le prix spécial du jury des «Acteurs du libre» pour son rôle crucial dans la défense des libertés informatiques.

[Génération NT] HDMI 2.1 bridé sur la Steam Machine: la faute à Linux!

✍ Mathieu M., le mardi 9 décembre 2025.

La nouvelle Steam Machine de Valve, bien que matériellement compatible avec le HDMI 2.1, sera limitée par son logiciel. En cause: SteamOS, basé sur Linux, ne peut intégrer les pilotes open source nécessaires, bloqués par le HDMI Forum. Valve doit donc trouver des parades, comme la compression de couleurs, pour offrir du 4K à 120 Hz, mais avec des compromis.

[leParisien.fr] Un bug informatique menace de bloquer le RER A et huit lignes de métro en 2038

✍ Candice Doussot, le mardi 9 décembre 2025.

Le tribunal administratif de Paris a condamné, jeudi 13 novembre, Alstom à réparer un défaut logiciel grave affectant le RER A, huit lignes de métro et six lignes de tramway. Sans correction, ces rames s’arrêteront le 19 janvier 2038.

[Les Numeriques] “Il est temps d'envisager sérieusement de quitter Windows”: l'appel de 20 associations pour passer sous Linux

✍ Corentin Bechade, le lundi 8 décembre 2025.

Fort de son succès récent sur l’allongement de la durée des mises à jour de Windows 10, le collectif «Non à la taxe Windows passe désormais la deuxième en faisant ouvertement campagne pour une bascule vers le logiciel libre.

Et aussi:

Commentaires : voir le flux Atom ouvrir dans le navigateur

  •  

Chez les fondeurs, TSMC reste intouchable, mais les acteurs chinois accélèrent au 3T 2025

TrendForce a dévoilé son classement des dix plus puissants fondeurs (par revenus) du troisième trimestre 2025. Il témoigne de la suprématie de TSMC ; le taïwanais domine largement le marché, et se paye aussi le luxe de signer l’une des plus fortes croissances trimestrielles (la troisième)... [Tout lire]
  •  
❌