Shared posts

12 Nov 13:27

LG Display’s 12-inch stretchable screen can extend by up to 50%

by Jean-Luc Aufranc (CNXSoft)
LG Stretchable display

A new category of devices – the stretchables – may soon be upon us, as after writing about Murata stretchable PCBs for medical applications a few days ago, I’ve now come across LG’s latest stretchable display that can be extended from its original 12-inch size up to 18 inches.

It’s the second stretchable display showcased by the Korean company, as the first prototype was unveiled in 2022 with an elongation rate of just 20%. The new display extends that to 50% and delivers a high resolution of 100ppi (pixels per inch) and full red, green, and blue (RGB) color (probably at its original size).

LG Stretchable display

LG Display further explains that their engineers have applied a number of new technologies “such as improving the properties of a special silicon material substrate used in contact lenses and developing a new wiring design structure to improve the panel’s stretchability and flexibility”. Durability was also enhanced by using a micro-LED light source of up to 40μm enabling the display to be repeatedly twisted and stretched over 10,000 times.

Such stretchable displays could be found on irregularly curved surfaces like clothing and skin, and the company expects it to be integrated into various products in the fashion, wearables, and mobility/automotive sectors. The company showcased several concepts for the stretchable display including an automotive touchscreen panel that stretches out into a convex shape and a wearable display attached to firefighters’ uniforms that provides real-time information.

LG Display stretchable screen

The company did not share many other technical details and it’s unclear when the new stretchable display will become available. All we have is a press release. I tried finding a few demos of the new 12-inch display, but all I could find was a 2-month-old video showing the first display prototype (that stretches by up to 20%) on clothing and bags at Seoul Fashion Week.

Via Liliputing

The post LG Display’s 12-inch stretchable screen can extend by up to 50% appeared first on CNX Software - Embedded Systems News.

12 Nov 12:46

Les meilleures innovations dans le secteur du BTP

by Carène M.

Dans le secteur de la construction, les tâches sont physiques, pénibles et répétitives. Ceci explique peut-être pourquoi la main-d'œuvre manque parfois. Pour pallier ces difficultés, les nouvelles technologies sont d'une grande aide et permettent d'économiser en temps, en argent et en efforts. Dans cet article, nous évoquons dans les grandes lignes les meilleures innovations matérielles dans le secteur de la construction.

Les machines écolo

La construction demande une forte ressource en énergie et en eau. Même si la solution zéro pollution est difficile à atteindre, aujourd'hui, des efforts sont faits dans ce secteur pour réduire au mieux l'impact environnemental. En choisissant la location matériel luxembourg de Boels, vous pourrez réduire les nuisances sur l'environnement. La gamme d'engins Eco permet de minimiser l'empreinte carbone sur les chantiers et de réduire le bruit et les vibrations au sol. Le recyclage de l'eau utilisée lors d'une construction et celui des déchet est aussi un effort nécessaire à réaliser.

Les matériaux innovants

On le sait, l'humidité dans une pièce est mal venue, d'où l'importance de l'achat ou de la location déshumidificateur à condensation ou par absorption. Cependant, l'humidité peut être nécessaire pour réparer le béton grâce à une méthode récente et naturelle. Il suffit d'injecter de petites capsules contenant des bactéries et des nutriments au cœur de la fissure à combler. Sous l'influence du soleil et de l'humidité, ces polymères vont proliférer et combler le trou. Autre matériau innovant : le verre photochromique qui devient plus opaque sous l'effet des rayons lumineux. Ainsi, l'éclairage d'une pièce pourra rester constant.

Les drônes

Ce sont loin d'être des gadgets dans le BTP ! Les drônes permettent d'atteindre des endroits inaccessibles sans prendre de risques pour cartographier un site et le modéliser afin de travailler sur une maquette numérique. Le drône permet aussi de gagner un temps précieux car il permet de pulvériser de la matière anti-rouille ou anti-mousse sans avoir à utiliser une nacelle. Il peut aussi relever des stocks en un seul vol et faire un diagnostic thermique, entre autres exemples.

Le BIM

Le BIM ou « Building Information Modeling » est une méthode numérique de gestion de projet en temps réel. C'est un plan en 3D auquel tous les acteurs de la construction vont pouvoir toucher et ce, de la phase de conception, jusqu'à la phase de maintenance. Ainsi, il sera possible de voir apparaître toutes les demandes du client, les matériaux disponibles chez le fournisseur, les changements à apporter, etc. C'est in fine une méthode de création de projets interactive et évolutive qui révolutionne le monde du BTP. Plus besoin de suivre les étapes rigides et restreintes entre le maître d'ouvrage et le bureau d'étude, comme autrefois.

La robotique de construction

La présence de robots dans le secteur du BTP a toujours été vue comme un levier positif. Ceci parce que le secteur est le plus concerné par les accidents de travail et que la main d'œuvre manque. Ainsi, la robotisation des tâches résout ces deux enjeux. De nombreux types de robots existent : les robots porteurs de charges, les lisseurs d'enduits, les robots pulvérisateurs pour peindre les murs ou le robot perforateur, etc. Récemment, une start-up australienne a créé un robot maçon capable de monter les murs d'une maison en 2 jours.

La réalité virtuelle

Nous l'avons vu, le BIM permet de construire en 3D et de façon interactive. La réalité virtuelle permet d'aller plus loin encore et de pénétrer au sein de la maquette, comme si l'ouvrage était réel. Cela permet à tous les acteurs du projet et au client de mieux percevoir le rendu final du projet, de mieux déceler les manques, de travailler en meilleure intelligence et plus rapidement.

Cet article Les meilleures innovations dans le secteur du BTP est apparu en premier sur OBJETCONNECTE.COM.

12 Nov 10:15

Vatican, Microsoft create AI-generated St. Peter's Basilica to allow virtual visits, log damage

The Vatican and Microsoft on Monday unveiled a digital twin of St. Peter's Basilica that uses artificial intelligence to explore one of the world's most important monument's while helping the Holy See manage visitor flows and identify conservation problems.
12 Nov 10:11

LG shows off first display that can stretch by 50%

by Simon Cohen
12 Nov 10:10

Pi Board : un plateau d’échec réel pour un adversaire virtuel

by Pierre Lecourt

Le Pi Board est une création de Tamerlan qui documente toute son approche autour de ce projet de jeu d’échec réel et intelligence virtuelle.

A vrai dire, j’ai déjà connu des produits commerciaux de ce genre, il y a fort longtemps j’ai croisé un plateau de jeu d’échecs de ce type. Un système d’aimants déplaçait les pièces pour les faire avancer sur toute la surface. Avec toutefois une petite nuance, le matériel ne savait pas vraiment où vous placiez vos propres pions et il fallait entrer pour chacun de vos déplacements les cordonnées de départ et d’arrivée. Ajoutez à cela un adversaire adoptant une petite dizaine de stratégies seulement, ce n’était pas vraiment une merveille. D’autant que le plateau était vendu très cher et que les pions étaient en plastique creux pour être assez légers pour se déplacer. Bref, ce n’était pas vraiment génial.

Ici, le Pi Board qui est présenté est beaucoup plus complet, il faut dire que les éléments qui entrent en jeu sont un peu plus récents. On retrouve un plateau de jeu classique avec en dessous un système de moteurs pas à pas sur deux axes X et Y qui déplacent un aimant sous les pièces pour les faire avancer. Si l’idée de base a été d’employer un électro aimant, celle-ci n’a pas été retenue. Faire monter ou descendre un aimant classique sous le plateau semble plus efficace que d’alimenter un électro aimant, c’est également moins gourmand en énergie et moins dissipateur de chaleur. Ainsi le dispositif avec un simple aimant qui vient se placer sous la pièce à déplacer peut suivre le damier du jeu plus efficacement et sans d’autres contraintes.

Une fois cette base technique développée, on pense évidemment aux solutions comme les graveuses laser qui déplacent une tête de gravure avec le même genre de mouvements sur deux axes, il a fallu optimiser. L’idée est d’avoir une réponse rapide à vos mouvements de jeu pour que la partie ne soit pas ennuyeuse. Si votre coup joué doit attendre le repositionnement de l’aimant sous la nouvelle pièce entre chaque étape, cela risque de ne pas vraiment donner envie de jouer. Un algorithme de positionnement a donc été développé pour que le Pi Board ne place pas l’aimant de déplacement à son point de base après chaque coup. La machine se souvient de sa place et peut calculer à partir d’une grille son prochain déplacement automatiquement. Cela fluidifie grandement la partie. 
D’autres éléments techniques ont également dû être optimisés comme les déplacements spécifiques du roi et de la reine ainsi que celui du cavalier. Des stratégies ont donc été mises en place pour gérer ces mouvements.

Enfin, le pilotage par une carte Raspberry Pi a permis de développer des solutions de jeu adaptables au niveau du joueur, capables de choisir les blancs ou les noirs mais également de relier le plateau du Pi Board au réseau pour piloter une partie avec un autre joueur. On imagine assez bien l’intérêt d’avoir ce genre de solutions déployées chez deux joueurs distants qui peuvent ainsi s’affronter « face à face ». Reste que pour le moment il faut entrer ses propres mouvements de jeu au travers d’un petit écran tactile situé sur le côté. Je ne sais pas comment la machine pourrait analyser le plateau de jeu. Peut être avec une caméra au dessus du plateau pour analyser les pièces et détecter les mouvements du joueur humain ? Peut être avec des capteurs NFC sous les pièces ?

Comme d’habitude, je suis assez admiratif du résultat obtenu et de l’ingéniosité de l’auteur. C’est une excellente idée d’apprentissage et de perfectionnement avec de nombreux défis à relever autant au niveau matériel pour gérer les pièces qu’en programmation pour en faire un jeu agréable.

Source : Tamerlan

Minimachines.net en partenariat avec Geekbuying.com
Pi Board : un plateau d’échec réel pour un adversaire virtuel © MiniMachines.net. 2024.

12 Nov 10:07

ChinAI #288: The Art of the Strategic Pregnancy

by Jeffrey Ding

Greetings from a world where…

people talk to each other like chickens speaking with ducks [鸡同鸭讲]

…As always, the searchable archive of all past issues is here. Please please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay support access for all AND compensation for awesome ChinAI contributors).

Feature Translation: People at large tech companies, strategically preparing for pregnancy

Context: In the past, in order to get promotions and raises, we didn't dare to have children, as we were afraid that getting pregnant and having children would affect our career development. Now, in order to keep our jobs, we are considering strategic pregnancies,” one poster comments on a social network. This week’s feature translation (original Chinese from Perpetual Light Studio[极昼工作室] covers this phenomenon of “strategic pregnancies” [战略性怀孕], which has gained momentum in the past two years. From the article:

Searching through a certain social platform dominated by female users (Xiaohongshu, I think), more than 25,000 notes are about “whether you will be laid off after a strategic pregnancy.” Someone posted a "practical guide" to share how she safely avoided three waves of layoffs in large companies from pregnancy to maternity leave, as her department went from more than a dozen people to only herself.

Key Passages: Our story begins with Yuanyi Ma, a 30-year old in her fifth year in the game development industry. In her prime, she wants to rise, but there are no vacancies in management positions, and her company has started to subtly scale back expenses and benefits (e.g., the cafeteria napkins have been replaced with cheaper brands).

  • Ma’s strategic pregnancy plan: 1) get pregnant in the second quarter of the year, 2) disclose her pregnancy to the firm only after she gets her bonus, 3) make sure the third trimester of her pregnancy coincides with the timing of the national and provincial civil service exams, 4) after giving birth, continue to review during maternity leave and participate in other exams in the first half of the following year, 5) finish breastfeeding and then catch up with the recruitment season for other positions.

  • The article continues: “She did it. She had pregnancy reactions in the first three months, tried to hold back when she wanted to vomit, and went to the bathroom quickly after meetings. When she announced her pregnancy, she was relieved. Another colleague who joined the company at the same time seemed to be inspired and went to skip rope every night after getting off work. Mai knew that skipping rope is conducive to ovulation.”

In mid-September of last year, Yue Du and her husband had only been married for two months when she received a layoff notice. Luckily, someone in management had tipped them off in advance, so the young couple decided to get pregnant as soon as possible to buy some time with human resources (HR).

  • Negotiations with HR delayed her firing, which allowed her to find another job at an investment fund company. Unfortunately, that company also struggled, so she had to turn to side hustles, such as audiobook reader and part-time auditor. The only one that stuck involved selling insurance policies and assisting doctors to help out-of-town patients.

  • Du’s husband noticed that she had become very tight about money. “He changed from buying ribs from Meituan to going to the vegetable market to compare prices.”

  • Some HR professionals advise against undertaking a strategic pregnancy to keep one’s job, as it only brings a short-term sense of job security. Upon return from maternity leave, female employees sometimes struggle to secure promotions and salary increases. The article cites a recent “Survey Report on Fertility Intentions of the Suitable-Age Population,” jointly published by statistics bureaus in various areas, which cites “inadequate implementation of maternity allowances, holidays, and employment guarantees” and “worries about the impact on personal career development” as two of the main obstacles to having children.

Ma continued to prepare for exams during and after her pregnancy. She took the national civil service exam two weeks before giving birth. After just two weeks of rest after a C-section, she continued her studies, doing math problems in bed while breastfeeding. During her six-month maternity leave, she took four exams and applied to posts at other companies.

  • One month after returning to work, she received an offer to become a teacher at a vocational school. After taking up the post, in order to better her chances of future promotions, Ma chose to be a head teacher, so she was busy every day with teaching plan and lesson preparations.

  • The article concludes: “One night at eleven o'clock, several girls in the class had a conflict in the dormitory, just as Ma’s daughter had a fever and had to go to the hospital. On the way to the hospital, she kept communicating with the students' parents, teachers on duty, class leaders, and the students involved. The phone was silent only after she got the medicine for her daughter. Lying in bed at one o'clock in the night, she turned on her phone and sent messages to the parents one by one to comfort them again. ‘No job is the best, they are all cities under siege.’ Ma sighed.”

For more stories of strategic pregnancies, see FULL TRANSLATION: People at large tech companies, strategically preparing for pregnancy

ChinAI Links (Four to Forward)

Must-read: Granta 169: China

The blurb for a special issue of Granta on China: At a time when China has become a unifying spectre of menace for Western governments, this issue of Granta seeks to bring the country’s literary culture into focus. I’m really looking forward to reading Han Zhang’s “Picun”, an essay about working class literature in a migrant worker’s colony outside Beijing.

Should-read: Epic’s overhaul of a flawed algorithm shows why AI oversight is a life-or-death issue

I’m about two years late to this article in STAT+ by Casey Ross, but I recently came across this citation in a working paper, and it is a very important read about the implementation lags of AI systems in medical systems. Let me know if you can’t access it, and I can share a pdf.

Should-read: Concordia AI Fall 2024 Hiring Announcement

Concordia AI is hiring for roles in their Beijing and Singapore offices: International AI Governance Research Manager/Researcher; Foundation Model Safety Research Engineer; Operations Manager/Associate. Check out the announcement page for more details.

Should-read: Beijing Academy of AI launches LLM debate platform

This article (in Chinese) introduces the FlagEval Debate platform which allows 2 LLMs to debate against each other based on prompts generated by the platform. H/t to Shao Heng for sharing.

Thank you for reading and engaging.

These are Jeff Ding's (sometimes) weekly translations of Chinese-language musings on AI and related topics. Jeff is an Assistant Professor of Political Science at George Washington University.

Check out the archive of all past issues here & please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay for a subscription will support access for all).

Also! Listen to narrations of the ChinAI Newsletter in podcast format here.

Any suggestions or feedback? Let me know at chinainewsletter@gmail.com or on Twitter at @jjding99

11 Nov 12:09

Palmshell SLiM X4L : un MiniPC Intel N100 signé Radxa

by Pierre Lecourt

Le Radxa Palmshell SLiM X4L propose un fonctionnement classique et une orientation vers des usages très spécialisés. La marque que l’on connait pour ses cartes de développement sous SoC ARM a bien compris l’intérêt de ce processeur Intel Alder Lake-N.

De nombreux utilisateurs de minimachines « classiques » sous Intel N100  avec une ventilation active se servent aujourd’hui de ces engins pour  des usages très spécialisés. Très loin de ce qu’imaginait Intel pour ces solutions pensées comme « l’après Celeron/Pentium » pour les particuliers. Ces minimachines servent à piloter des outils réseaux, créer des serveurs, gérer des imprimantes 3D ou autres outils du genre. Avec des cartes de développement dont le prix est monté en flèche et dont l’équipement annexe n’a pas baissé et la baisse du tarif de ces MiniPC « prêts à l’emploi », il est désormais souvent plus intéressant de glisser un engin complet, enfermé dans un boitier, sous Intel N100 et bien accompagné de mémoire et de stockage rapide que de rassembler tous les éléments d’une solution de développement type Raspberry Pi pour certains usages.

Radxa l’a compris et a décidé de proposer le Palmshell SLiM X4L comme alternative à ses cartes de développement classiques. Annoncé à partir 129.99€ HT, c’est un engin qui devrait proposer un excellent ratio performances et encombrement pour des utilisations spécialisées. On retrouve donc ici un Intel N100  enfermé dans un boitier tout en largeur. Pensé pour être positionné en VESA facilement, il dispose sa connectique sur deux côtés uniquement.

On retrouve ainsi pas moins de quatre USB 3.2 Type-A, un Ethernet 2.5 Gigabit, deux sorties HDMI 2.0, un jack audio combo 3.5 mm et un Antivol type Kensington Lock. L’alimentation est assurée par un port USB Type-C Power Delivery 3.0 qui demandera du 12V et 3A… Mais à vous de le fournir, l’alimentation n’est pas prévue dans la machine. Un module Wi-Fi6E et Bluetooth 5.2 est également intégré.

Autour du processeur Intel N100, on retrouve de 8 à 32 Go de mémoire vive LPDDR5-4800 et 32 Go de eMMC de base associé à 128 Go NVMe PCIe 3.0 monté en M.2 2280. Ne vous laissez pas impressionner par son châssis qui fait penser à une solution passive, il est bel et bien équipé d’une ventilation active. 

Je ne suis pas certain que cette offre soit la plus pertinente pour un particulier à la recherche d’un MiniPC pour des usages moins orthodoxes que celui d’un engin de bureau. Si la place ne compte pas à ce point là et qu’un boitier un peu plus épais vous convient, il y a une foule d’autres solutions plus accessibles et peut être plus intéressantes niveau tarif et évolution que ce modèle. Pour un usage vraiment spécialisé comme de l’intégration derrière des écrans ou dans des kiosques, ce design peut par contre avoir du sens. Et si vous êtes vraiment vraiment à la recherche d’une solution la plus compacte possible, Radxa propose également la X4, une carte de développement sous Intel N100.

Radxa X4 : la carte de développement sous Intel N100 est dispo

Sources : Bret.dk via Liliputing

Minimachines.net en partenariat avec Geekbuying.com
Palmshell SLiM X4L : un MiniPC Intel N100 signé Radxa © MiniMachines.net. 2024.

11 Nov 12:09

Meta should improve its Store to support more its developers

by Skarredghost

There is a bit of turmoil in the VR developer communities because developers are experiencing more difficulties in making their games noticed on Quest. They are all asking Meta to take action, otherwise they risk going out of business. Let me tell you the main complaints, some possible solutions, and why it is important to speak about this.

The merging of App Lab

The problem

I’ve always hated the curation of the Meta Store that relegated us indie developers to the secondary “App Lab” store. I’ve always thought that made no sense to divide developers into Class A and Class B developers and that everyone should have the same opportunity to publish his own game and make it visible to the community, hoping that it can become a hit. It’s fair this way. So when App Lab was eliminated and all the games were merged into the main (Quest) Horizon Store, I was very happy.

But the problem is that the Horizon Store was probably not ready for this merge. App Lab contained both interesting indie games and cheap experiments. After the operation, many people complained about the fact that their home page of the store was just full of shovelware, and others that they were suggested mediocre content. I remember one guy saying that he was suggested to play “Orilla Tag”, a clear rip-off of Gorilla Tag (I wonder if there is also a “Rilla Tag”, “Illa Tag”, and so on, until the most infamous of them all, that is called just “G”). This means that the system that is suggesting to people the game to play is not working very well, and is giving visibility to crappy content that no one wants to play instead of rewarding little gems made by indies.

Together with the suggestion engine, there is also another big problem in the store, which is the Search Engine. Finding games using a part of the name is not always easy: sometimes the wanted game does not appear, or it is not the first result, and so on. This means that if someone wants to find a game that saw a couple of days ago in a Youtube video and so he only remembers some vague resemblance to the name, he may not find it. And these are all lost sales for the developer of that game.

meta store looking for
If I try to look for “Hitmotion”, I don’t find my game “Hitmotion: Reloaded”

A possible solution

There is one simple solution for the search functionality: Meta should fix it. Searching the web is not a mysterious topic anymore, and there are many algorithms for that. Considering Meta’s budget, I imagine it is just a matter of making this a priority and taking the time to fix it.

As for the suggestion engine, it is always been a problem since the Gear VR times, when people started receiving nonsense notifications like “Since you liked Arizona Sunshine, go playing Hello Kitty Word 2000” and published them on Reddit so that we all could have a laugh. I think it’s time to fix it once and for all: in this case, it is not so trivial as fixing the search functionality, though.

Makes totally sense, thanks Meta (Image by Reddit user quiken4)

There is not a single standard for suggesting something to users, but there are different strategies out there. And there are also different characteristics through which you can choose the content to suggest: for instance, what has more weight, the type of content (e.g. action vs adventure), its length (short vs long), the reviews it had, or the previous history of the user? The algorithm has to choose how to mix all of these things to give the user a good set of suggestions. It’s not easy, but a company like Meta, which is also one of the leaders in the AI sector (think about the Llama models…), should be able to carry on this task.

These are not the best suggestions in the “Free apps” category…

But there is even a bigger problem, that is how to create a good mix of certified quality games and hidden gems. One naive approach would be to allow all games on the store, but give visibility (by ranking them higher in the searches or putting them in the Featured section) only to the ones by teams that are certified to make good games. But this would de-facto shadow ban all the indie games from new small teams, creating a situation very similar to the one we had before with App Lab. The solution would be in my opinion to create a mix between the two: on the front page of the store that is customized for every user, the system should provide most suggestions that are known games (e.g. Batman: Arkham Shadow, Gorilla Tag, Max Mustard) and then leave a few slots to unknown games that may be hidden gems that the user may like. Then the system should track how many of these suggested unknown games are actually getting good play time and positive reviews, and the more they perform well in terms of engagement, the more they should be promoted up in the rankings. The ones that are played and mostly disregarded, instead, should be given always less visibility, even if they come from well-established companies. I believe this would be a fair treatment because games would be the facto evaluated by their users: the more a title is liked by the Quest users, the more would become a hit worth being showcased.

The visibility of Horizon Worlds

The problem

It’s been since ages that Meta has tried to make a successful Social VR application. Its last one is Horizon Worlds, which was marketed as the ultimate social VR experience and launched with much hype, but that never had massive success. Competitors like VRChat and Rec Room always had more users and the press mocked Facebook, a social company, for not being able to create a compelling social VR experience. Since Meta invested a lot of money in Horizon and absolutely wants to dominate the social XR space, because social brings lots of things with it (including the data of people), it started to shovel into our throats Horizon Worlds everywhere. If you open the Meta Horizon app on your phone, you will notice that the home page is all about Horizon Worlds worlds and to open the actual store of apps, you have to click on a tiny button in the upper-left corner or scroll down the page for a while. When you launch your Quest, the first tab that opens is the “Horizon Feed”, which suggests a mix of games, Horizon Worlds worlds, and videos from Youtube creators (and I have never asked for the last two things). When you search for something in the Store, the Store tells you both games and Horizon Store destinations with that name. Sometimes even the game suggestions are not about games but about Horizon Worlds places. At the latest Meta Connect, Meta proposed a re-design of the Quest runtime, and guess what? There was an easy way to go to Horizon Worlds portals…

home of meta horizon mobile app
This is what I see when I open the Meta Horizon app on my smartphone: those are all Horizon Worlds worlds

I understand that there are some managers in Meta that have to reach their annual goals and so they are doing everything to have a certain number of users in Horizon by the end of the year. But this is creating issues for everyone. As a user, I find the mobile app totally unusable: it should make it easy for me to buy games and configure my Quest, while it makes it easy to go to Horizon Worlds or to configure my Meta Avatar (that is mostly useful for Horizon Worlds). As a developer, I find it frustrating that my content has to compete not only with the other games on the store but also with Horizon Worlds destinations, which are even given priority on mobile. Someone is also complaining that some people prefer a free mediocre Horizon World ripoff destination to the original paid game on the Meta Store. Again, this is hampering the sales of content on Meta Store.

I would also add that I wonder if there can also be legal implications attached to what Meta is doing. I’m not a lawyer, but I guess is not very fair that Meta is showcasing so much of the content of its own social VR application when there are other social VR applications on its store (e.g. VRChat) that are not getting the same treatment. The recent Digital Markets Act in the EU puts more pressure on the so-called “gatekeepers” to not favor their own products in the search results of their platforms. So, if I were Meta, I would be careful and evaluate what I am doing also on a legal standpoint before some antitrust entity wakes up and notices these potential issues.

A possible solution

I’m sure the skibidi world is amazing, but still…

I would re-design the whole experience on the mobile Meta Horizon app, making it more usable and giving more visibility to content on the store. And I would do the same on all platforms, from the web store to the runtime of Quest. I would also remove all the “Horizon Worlds” results from the search engine of the Store: if someone wants to look for a Horizon Worlds world, he had better open the Horizon Worlds app. Otherwise, it is like if I opened the Start menu of my Windows PC, and when I start typing “Wo” instead of finding “Microsoft Word”, the first suggestion would be the “Wololo Potato” world in Roblox. Wouldn’t this feel strange?

quest companion app meme
Happens every time…

If Meta wants still to drag more people into social VR, it could create in its interface a little “Featured Worlds” section, where it hand-picks 4-5 spotlight experiences across all social VR worlds (Horizon Worlds, Rec Room, VRChat, …) that are absolutely worth a look. This could have value for both the users (who can so discover majestic social VR experiences) and developers of social VR content (because they have the opportunity to be spotlighted), but wouldn’t hurt the visibility of Quest apps on the main Store of the headset.

Furthermore, I think that Meta should also realize that probably Horizon Worlds is not what the VR community is looking for at this moment. If you need to always brute-force your product in front of the eyes of the people to gain new users, then it means that the product does not fit with the needs of the people. Consider also that notwithstanding all this visibility, Horizon Worlds is not even the most popular social VR world on the market: this says a lot. Google Chrome has been installed on every Android phone and it immediately became the to-go browser for everyone paired with Google Search because both products were crazy good before the recent enshittification. If this is not happening with Horizon Worlds, I think Meta should question itself if it is really on the right path with this product.

The Store Cut

meta quest horizon feed
The Horizon Feed and Horizon Store inside my headset

The problem

Meta has still a 30% cut on all the transactions happening in the Store. 30% is really a lot of money, especially for small teams that are trying to make their business sustainable.

A possible solution

I know that managing a store is complicated and expensive. And I also know that Meta is subsidizing its cheap hardware with the money it earns from its store. So I’m not asking Meta to remove the Store fee, because that would make no sense.

But I still think there are ways to find a middle ground with the needs of the developers. For instance, the store fee may be reduced for the first period of “start up” of the various products or companies. Or it could be reduced for the products with sales below a certain threshold.

Why Meta’s support is important

meta quest 3s front
Front side of my Meta Quest 3S

At this point, you may wonder why Meta should care about supporting small developers when it can afford to have big titles like Asgard’s Wrath 2 or Batman: Arkham Shadow that are self-funded. The reason is that the big titles are important, but it is the vibrant indie ecosystem that is keeping the platform alive, with its passion, innovation, and new ideas. I repeat it every time: many VR big hits have been developed by small indie studios. Two huge examples of this are Beat Saber and Gorilla Tag. At this moment, Meta can have a successful platform only as long as its developers are happy with it. And “happy” means that after they invested a lot of time and money to develop a piece of content, they can find enough people on the Meta Horizon Store that play their content and pay for it, making their business sustainable. Also because if we pay 30% of cut to the Store, the Store should give us a good service in return. If this doesn’t happen, all the developers will leave the platform, leaving Quest with a poor ecosystem and very bad press.

You may then wonder where could we go, considering that there is no platform as big as Quest now. Well, first of all, “now” is the key word here. Wait a few years and the Apple and Google stores would have become more relevant in XR, and if they offered better opportunities to devs, all people could jump ships. Then, remember that there are many ways to make money in XR, and enterprise XR is one of them: we are not forced to make consumer content. Or we could also simply do other jobs, like doing standard 3D content (which has an even bigger audience) or jump on the AI bandwagon. Actually, many friends have already left VR and are suggesting me to do the same and look for money elsewhere. We developers have many options, if we stay in VR is just for our passion.

It’s good to see that someone at Meta is already listening to our complaints

Don’t misunderstand the tone of this article, though: I’m grateful for what Meta is doing for XR, for having released the Quest headset, and for having so created a VR ecosystem that is big enough that is sustainable for some developers. But I think that if we want to make XR succeed, we need to do this all together, and we should all support each other for this to happen. And this includes Meta in doing its part in helping us in doing our job. Even if this means making some present sacrifices like not putting its own product under the spotlight, because this can bring many long-term advantages for us and also for Meta itself. I really hope this can happen.

(Header image by Meta)

The post Meta should improve its Store to support more its developers appeared first on The Ghost Howls.

11 Nov 10:15

TSMC To Halt Shipments to China

by Brian Boyle

Whoops, how’d that slip through?

US sanctions have been designed to keep high-powered computer chips from entering China, but last month one of TSMC’s chips was discovered inside a processor made by the Shenzhen-based Huawei. Now, according to a Reuters report this weekend, the US government has ordered the chipmaker to halt orders of advanced chips to the country starting Monday.

TSMC You Later, China

US rules do not ban TSMC from selling all chips to China, just those that are high-powered and could be used for advanced artificial intelligence purposes. But while imposing export controls on high-powered chips is easy, enforcing them is much harder. That’s because TSMC makes chips based on customer designs, and experts say that chips are so complex it can be difficult to ascertain their power based on design alone. 

In October, a teardown of a Huawei AI processor by research firm Tech Insights found one of TSMC’s high-powered chips. TSMC then flagged the discovery to the US Commerce Department, claiming it was unsure how the chip made its way into the China-made hardware. The incident has raised enough of a panic to spark even more stringent export controls:

  • Following the notification, the US Commerce Department ordered TSMC to impose even more beefed up restrictions on certain chips, a source told Reuters. TSMC has told Chinese clients it is suspending shipments starting Monday, and will introduce a tighter review process on designs likely to involve the US government, sources told the Financial Times.
  • That likely means an even bigger dent to its business. The share of TSMC’s revenue from China has fallen from around 20% in 2019 to 12% last year amid the increased export controls (the company still generated around $9.8 billion in revenue in October alone, it recently announced).

A New Era: For its part, TSMC in a statement insisted it was a “law-abiding company.” The Taipei-based firm operates in accordance with US rules to maintain access to critical US tech. But with a new administration en route that has criticized Taiwan’s chip industry, TSMC’s continued seat at the US table looks slightly in doubt. One source told the FT that TSMC’s cooperation is “designed to underscore that we are the good guys and not acting against US interests.”

The post TSMC To Halt Shipments to China appeared first on The Daily Upside.

11 Nov 10:12

📡 Effet papillon: tout un Univers dans un Trou noir ⚫️ ?

by Adrien BERNARD
Une petite vague qui provoque un raz-de-marée. Dans l'Univers primordial, de minuscules fluctuations quantiques pourraient avoir eu un effet profond sur le cosmos. À l'aide de simulations...
10 Nov 16:44

Wall Street Does DeFi

by Sean Allocca

DeFi is in the middle of an identity crisis. Born out of the 2008 financial meltdown, the decentralized finance movement aimed to reinvent a too-big-to-fail global financial system that wasn’t always looking out for the little guys. Why give money to the big banks when you could create a peer-to-peer system you controlled? 

But if cutting out Wall Street with blockchain technology was an objective, the “cypherpunks” have failed: the world’s largest banks are about to cash in on the technology they’ve been championing for some two decades. BlackRock, WisdomTree, and a handful of others have all launched new digital-native funds in recent months, from tokenized money market funds to digital Treasuries. These new products are much cheaper to operate and come with additional benefits, like the ability to instantaneously settle transactions. They’re also tapping into an industry that is transacting trillions of dollars on-chain per month. 

“It’s potentially one of the biggest technological advances in markets since trading became electronic,” said Aaron Kaplan, CEO of the digitally native brokerage Prometheum, the first firm to get the regulatory green-light to safeguard and clear digital assets. “It has the potential to disrupt not just the securities industry, but much larger financial institutions.”

Chain Wallets?

The beauty of the underlying technology is that it can “tokenize” real-world assets, anything from real estate, to stocks, to fine art. The tokens are held on a blockchain, a digital ledger, and can be bought and sold just like any other security. Whoever owns the token owns the real-world asset. The size of the market for all this is mind-boggling: the total value of tokenized assets could reach $24 trillion by 2027, according to a report by the World Economic Forum, which would account for nearly 10% of global GDP. 

“DeFi has these interesting use cases, but it’s essentially stuck in its own universe,” said Kaplan, who was called to testify before a Congressional committee last year. Blockchain is a pseudo gray area for regulators as assets like cryptocurrency have yet to be defined as a commodity or a security. “The innovations have already been incubated in the DeFi space,” he said. “Now, they’re trying to figure out how to use them for traditional securities.” 

The biggest nut for companies to crack is securing these digital assets and keeping them out of the hands of cybercriminals. Regulators are lobbing lawsuits at blockchain-based exchanges, and politicians are calling for more scrutiny before the tech transforms how assets are traded and secured. 

Stuck in the BUIDL. BlackRock became one of the first major firms to launch a digital asset fund in March, and it’s paying off. Its Institutional Digital Liquidity Fund (BUIDL) now has about $500 million in assets under management. It effectively acts like a digital currency, offering a stable value of $1 per token that institutional investors can trade and collateralize in other on-chain transactions. The fund invests in cash, Treasury bills, and repurchase agreements, earning yield for holders. It’s become an early success story and a major test case for Wall Street’s bet on DeFi.

Being among the first to market has allowed BlackRock, the world’s largest asset manager, to build BUIDL into one of the largest and most recognizable products in its class. But, with investment minimums of upwards of $5 million, these new digital funds aren’t going to be a mass market any time soon. But for the super rich, it’s a no-brainer. 

Block & Tackle. A key advantage is the ability to instantaneously settle transactions. Take traditional yield-earning funds like bonds, for example. After the transition to T-1 in the US, fixed income products generally take one day to settle, allowing time for the assets to be cleared and delivered to the seller. But, on-chain funds close instantaneously, meaning the funds can be returned and redeployed right away, saving hours of wasted time. 

When you’re dealing with millions of dollars, every second counts. 

Faster and Cheaper? For blockchain proponents, this is just the next step in trading technology. Once upon a time, traders held paper tickets on the exchange floors as proof of ownership of an asset. When the stock market started hitting hundreds of thousands of trades per day, brokers needed a better way to record trades other than on glorified Post-It notes. With the advent of online record keepers in the 1970s, ticket receipts went digital.

Kaplan and other proponents expect blockchain will be the greatest leap forward since that switch from paper to digital. “It’s just inefficient the way it’s being done now,” he said. 

The post Wall Street Does DeFi appeared first on The Daily Upside.

10 Nov 10:54

From vision correction to smart tool – how the contact lens market has advanced

by Staff Writer

In the UK, figures from the Association of Contact Lens Manufacturers (ACLM) highlight that the contact lens market has grown considerably. According to the report, the size of the British contact lens market reached £360 million in 2023. All in all, the UK market now represents the largest of all the countries in the survey, accounting for around 19% of contact lens wearers. 

Per Simon Rodwell, Secretary General of the ACLM, this surge can be attributed to continuous innovations in lens technology, which have enhanced comfort, convenience, and functionality, strengthening their appeal to a broader audience. The growing popularity of contact lenses in the UK reflects a wider trend where these vision aids are no longer viewed merely as alternatives to glasses but as sophisticated devices capable of much more.

In this article, we'll look at how contact lenses have undergone significant changes, exploring their modest beginnings and current medical-technological applications.

The humble beginnings of contact lenses

The journey of contact lenses from basic vision correction aids to advanced wearable devices is both fascinating and inspiring.

Its origins trace back to the pioneering work of Adolf Eugene Fick, a German ophthalmologist who is credited as the inventor of the first successful scleral contact lens. Given the limited availability of suitable contact lens materials, however, Fick was unsatisfied with the discomfort of blown glass shell contacts that could only be worn for up to two hours at a time. 

Nevertheless, it was Fick's work that effectively laid the groundwork for all future contact lens designs. By the 1930s, the development of thermoplastics signified a massive step forward for scleral contact lenses, contributing to their popularity as an alternative to eyeglasses.

Technological advancements continued well into the 1970s when soft contact lenses were finally given US FDA approval and embraced by the general public. 

Soft contacts dominate the modern market

Today's contact lenses are far more advanced than those glass-blown prototypes by Fick.

Notably, the market offers a wide range of options tailored to various needs and preferences, including daily disposables, extended-wear lenses, and specialised lenses for conditions like astigmatism and presbyopia. Moreover, advancements in materials have made lenses more comfortable and breathable, reducing the risk of eye infections and irritation. 

The accessibility of these advanced contact lenses has also improved. No longer limited to in-store shopping, consumers can now purchase contact lenses online. Established retailer Vision Direct, for one, allows customers to purchase a variety of lenses such as multifocal, toric, and even coloured contacts. This platform delivers affordable contact lenses straight to customers' doorsteps, making vision correction more convenient than ever. 

Coupled with the introduction of digitalised eye tests that can conduct eye exams in as little as five minutes, the shift towards online retail has democratised access to high-quality eye care services and products, further enhancing the usability and appeal of contacts.  

Smart contact lenses and the future

One of the most promising applications of smart contact lenses was developed by wearable technology firm InWith.

First unveiled during the 2022 Consumer Electronics Show in Las Vegas, the startup claims that the smart contact lenses can display real-time information and offer wearers "tunable vision" that can be controlled with a smartphone. The product announcement claims several hundred patents, which allow an augmented reality (AR) display chip to be used with and embedded into existing hydrogel contact lenses.

Smart contact lenses are also being explored for their potential to detect glaucoma early. Researchers from the Institute of Advanced Structure Technology at the Beijing Institute of Technology have developed a prototype designed to measure eye pressure accurately.

These measurements are typically taken during eye exams where a slight elevation in pressure indicates fluid buildup around the cornea, leading to a glaucoma diagnosis. With the dual-circuit “smart” lens design, however, such measurements can be monitored remotely for easier diagnosis, even in a wide range of temperatures. 

The UK contact lens market’s rapid growth reflects a broader trend: contact lenses are no longer just corrective devices; they are becoming essential tools with applications that range from healthcare to augmented reality. This technological evolution, coupled with the accessibility provided by online retailers, ensures that contact lenses will remain at the forefront of both fashion and medical innovation.

10 Nov 10:50

Un flash qui en dit long

by Marie Dollé

Rien capturer, tout révéler

Imaginez un appareil photo qui ne capture rien. Pas une image, pas un souvenir figé. Juste un rituel. Cet objet existe : le Polaroid LEGO. Lancé en début d’année, il ne promet rien d’autre que le geste. 

Un simulacre ? Non. Une leçon.

Nous vivons à l’ère des clichés éphémères, des souvenirs consommés avant même d’être vécus. Des images sans histoire, des preuves sans mémoire, un flot d’instants qui s’évanouissent aussitôt saisis. Dans ce monde où l’on veut tout immortaliser sans jamais savoir quoi retenir, ce Polaroid nous invite à réfléchir. Il célèbre l’acte de photographier sans céder à la possession. Lever l’appareil, cadrer… et renoncer. 

Polaroid LEGO : parce que rien, c’est déjà beaucoup

S’affranchir de la nowstalgie

Ce paradoxe résonne profondément à notre époque, où l’image règne en maître. Flashback, un projet lancé sur Kickstarter, prolonge cette réflexion. Avec lui, pas d’écran, pas de filtre : on capture, puis on attend. L’image n’apparaît qu’après 24 heures, laissant le temps de vivre l’instant. Une manière de renouer avec l’authenticité à l’ère du “tout, tout de suite”.

Cette quête répond aussi à un phénomène bien actuel : la nowstalgie. Nous cherchons à conserver le présent à tout prix, mais dans cette obsession, nous passons à côté de l’essentiel. En cadrant, en filmant, nous vivons déjà dans la nostalgie d’un moment que nous ne prenons pas le temps de ressentir. Flashback, ou encore le Polaroid LEGO, nous invitent à faire une pause. 

À lâcher cette compulsion de tout enregistrer pour mieux vivre l’instant.

Parfaitement imparfaits 

Dans cet esprit, Nonna’s Cam, imaginée par les créatifs de l’agence LOLA MullenLowe Madrid, dépasse le simple acte de photographier. Avec son doigt hyper-réaliste placé devant l’objectif, cette caméra argentique évoque avec tendresse - et humour - les maladresses de nos grands-parents, transformant chaque cliché partiellement obstrué en un hommage attendrissant aux souvenirs. 

Nonna’s Cam : un doigt, une légende

Ce projet invite à ralentir et à repenser notre obsession pour la perfection technique, si souvent recherchée via nos smartphones, et à redécouvrir la beauté de l’imperfection. Car finalement, la valeur d’une image ne réside-t-elle pas davantage dans l’émotion qu’elle fait naître que dans sa précision irréprochable ?

Quand la réalité vacille

Mais cette quête d’essentiel soulève une autre question : notre rapport au réel. Dans un monde où les deepfakes, filtres et réalités virtuelles brouillent les frontières entre vérité et artifice, peut-on encore croire aux images ? Peut-on distinguer le vrai du faux ?

Le concept de liar’s dividend, cette idée que le doute profite à ceux qui manipulent la vérité, aggrave la situation : dans un monde où tout semble falsifiable, même l’authentique peut être rejeté sous prétexte que "tout est peut-être truqué".

Face à ce doute généralisé, des alternatives émergent. La caméra ROC, presque comme un gardien du réel, promet des « moments incontestablement authentiques ». Son secret ? En combinant des capteurs, des preuves zéro-connaissance côté client sur l’appareil, et un environnement TEE inviolable pour attester des données des capteurs, elle garantit l’intégrité de ses images, avec pour objectif de restaurer la confiance dans ce que nous voyons.

ROC : la vérité, toute la vérité, rien que la vérité…

Nietzsche le pensait : les vérités absolues enferment l’esprit, là où l’incertitude libère. En rendant l’image irréfutable, ne risquons-nous pas de sacrifier ce flou poétique, cette part de mystère qui rend l’ordinaire extraordinaire ?

Réinventer la photographie

L’évolution de la photographie ne s’arrête pas là. De nouveaux concepts repoussent les limites de l’imaginaire et réinterprètent notre rapport à l’instant.

Prenez Soft une application qui capte le rythme de votre environnement pour transformer chaque photo en une œuvre vivante. Couleurs, saturation, grain : chaque aspect d’une image est influencé par l’atmosphère ambiante, inscrivant dans le cliché la respiration même de l’instant. Ici, il ne s’agit plus de reproduire fidèlement une scène, mais de capturer l’essence de ce qui l’entoure.

D'autres s'aventurent dans des contrées plus mystérieuses, comme Paragraphica, cet appareil singulier sans objectif ni capteur. Guidé par l’intelligence artificielle, il capte le murmure des lieux à travers leurs données invisibles, les traduit en mots, puis en images, révélant les secrets enfouis du réel. Ou encore DreamGenerator, qui mêle photographie classique et prompts créatifs, métamorphosant un simple cliché en une scène féerique ou rétro selon l’imaginaire de son utilisateur.

Paragraphica : la caméra qui voit avec des mots

Dans un élan encore plus symbolique, la Poetry caméra s’aventure ailleurs : au lieu d’une photo, elle offre un poème. Une manière d’embrasser l’instant en profondeur, de lui donner des couches de sens qui se dévoilent lentement, comme un souvenir qui se révèle avec le temps. Une belle métaphore, n’est-ce pas ? 

Photographier l’âme

Que faut-il retenir de tout cela ? L’évolution de la photographie, des premiers daguerréotypes à ces caméras qui réécrivent le réel, semble nous raconter une quête incessante : celle de capturer non seulement l’instant, mais aussi l’essence de ce qu’il contient. Sa capacité à suggérer plus qu’elle ne montre, à laisser place à l’imaginaire, au mystère.

Une chronologie que j'ai conçue, illustrant l'évolution de la technique vers le concept, où l'image devient une expérience à habiter autrement.

Après l’impressionnisme, l’expressionnisme et le surréalisme, voici venu le temps du selfpressionnisme : on ne cherche plus à exprimer le monde visible, mais à explorer notre expérience intérieure. Des appareils tels que la Poetry caméra ou le DreamGenerator transcendent leur fonction première : ils ne se contentent pas de figer le monde visible, ils invitent à l’habiter autrement. Ces objets, à la frontière du rituel et de la réflexion, nous incitent à ressentir une profondeur insoupçonnée dans l’éphémère.

Car oui, il y a des photos qui prennent tout. Et il y a celles qui laissent tout : le silence, le souffle, l’éclat fugace d’un instant. Une mémoire sans image, mais une lumière qui reste.

La photographie n’est pas une cage pour le temps, mais une clé. Une clé qui libère l’instant, le laisse vibrer, résonner. Alors, demain, quand vous cadrerez un moment, posez-vous cette question : capturer le visible… ou libérer l’invisible ?

MD 

10 Nov 00:21

The paradise paradox

The Paradise Paradox is the tempting but misguided belief that moving to a "paradise" will magically make everything better. It's the idea that a change of scene to somewhere beautiful or exciting will somehow sweep away all our problems. And while a fresh setting can give us a lift, the reality is often more complicated: paradise, as wonderful as it might be, won't always solve the everyday challenges we bring along with us.

Idealizing Paradise

Moving to your paradise might feel incredible, and it's a life-changer for some people, but it's easy to idealize paradise, imagining it as a place where worries simply disappear. Yet even the most stunning places could start to feel ordinary, and your conceptualised paradise may have its own challenges.

For instance, perhaps your paradise is a beautiful, distant, exotic land. Such a paradise, however, has the drawback of no longer being distant or exotic once you arrive and the day-to-day takes over. Someone living in a fast-paced metropolis might dream of calm and quiet, while someone surrounded by calm and quiet may dream of the rush and excitement of the big city.

Alastair Humphreys, who introduced me to the Paradise Paradox in his book Local, gave a great example in our podcast about Microadventures. In an interview on an American podcast, the interviewers couldn't imagine exploring Kansas to be exciting—it was known or even home to them. But to Alastair, a visit to Kansas would be an adventure, full of new things to see and do. It's a reminder that our ideas of paradise are often fueled by novelty—and novelty, by nature, doesn't last forever.

Expectations, the day-to-day, and Adaptability

Some of the Paradise Paradox may be explained by how easy it is to build up expectations. If we imagine paradise as the ultimate fix, it could lead to disappointment if it doesn't live up to the dream. As for vorfreude—the joy of anticipation—the build-up can sometimes overshadow the reality.

Then there's the simple truth that day-to-day life in paradise isn't a vacation. A tropical beach holiday lets you unwind and enjoy the scenery without a care. But living there might mean dealing with rainy seasons, work deadlines, or just regular old laundry.

Humans are also remarkably adaptable. We're good at adjusting to new surroundings, which means something initially remarkable may soon become the new normal. It's a bit like the grass-is-greener mentality we see in many areas of life. If you're a consultant, you might long for the stability of an in-house role, while in-house folks might crave the variety of consulting.

The Dark Side of the Paradise Paradox

In some cases, the Paradise Paradox takes on a more serious tone. The documentary film The Paradise Paradox, executive produced by Olympic ski racer Bode Miller and filmmaker Brett Rapkin, explores how people drawn to the beauty and tranquility of US mountain towns sometimes face mental health struggles. These picturesque communities, though visually idyllic, have experienced higher-than-average rates of suicide, highlighting the challenges of living in remote, isolated settings.

Strategies to Minimize the Pull of the Paradise Paradox

To reduce the effect of the Paradise Paradox, it helps to focus on what brings fulfillment in any setting, allowing us to appreciate where we are without feeling the constant pull of an imagined paradise. Here are a few ways to lessen the allure:

1. Cultivating Gratitude

Practicing gratitude allows us to recognize the value of what we already have. Focusing on the positives around us may make us less inclined to search elsewhere for happiness, realizing that contentment often lies closer than we think.

2. Prioritizing Experiences Over Material Goods

Research consistently shows that experiences bring longer-lasting happiness than possessions. Prioritizing meaningful activities and memories can bring joy and satisfaction, outlasting the initial thrill of accumulating more.

3. Simplifying Life

A minimalist approach can help reduce the constant desire for more—whether it's things, places, or achievements. Simplifying our lives can allow us to focus on what truly matters and bring a greater sense of peace.

4. Pursuing Meaningful Goals

Focusing on personal growth, nurturing relationships, and contributing to our communities can offer a lasting sense of fulfillment that doesn't depend on our surroundings. Meaningful goals and connections often provide the depth of happiness that location alone can't bring.

Paradise can certainly refresh us, but true and lasting happiness is often about embracing where we already are and what we already have.

Since writing this I keep seeing related quotes:

The only Zen you can find on the tops of mountains is the Zen you bring up there.

— Robert M. Pirsig (via Waking Up)

“We tend to attribute happiness to our circumstances (what happens to us) but studies suggest our happiness is really dependent on what we pay attention to.”

— Cal Newport, Deep Work (via Mind Over Tech)

"Happiness isn't getting what you want, it's wanting what you got."

— Garth Brooks

Wherever you go, there you are.

Also see:

09 Nov 16:34

Scala talks in-store retail media and how it's so much more than digital out of home (DOOH)

by Staff Writer

By Darren Cremins, Senior Sales Director - Retail Media and Connected Store, UK and Ireland, Scala

The evolution of retail marketing has taken on many forms, but one of the most dynamic developments is the rise of in-store retail media.

Some might see it as a glorified version of digital out of home (DOOH) advertising - displaying flashy messages on screens within stores - but in-store retail media is far more sophisticated, driven by data, customer interaction, and technology. It’s changing the way brands connect with consumers at the most critical point: when they’re about to make a purchase.

To simply define in-store retail media as DOOH is to ignore its vast potential for personalisation, targeted engagement, and sales optimisation. As we dive into what makes in-store retail media unique, we will discover why it stands leagues beyond traditional DOOH advertising, reshaping the future of retail.

The power of personalisation

One of the most striking differences between in-store retail media and traditional DOOH is the ability to personalise content to a greater degree. In-store media can draw on real-time data to target individuals or specific customer segments as they browse.

Imagine a shopper moving through the aisles, and the screen beside them highlights a product based on their previous purchases, loyalty card preferences, or even live stock levels in the store. This level of personalisation is impossible with DOOH installations that simply push a message to a broad audience.

The ability to integrate with customer loyalty programmes or mobile apps makes this even more powerful. When brands and retailers can understand not just general foot traffic, but who the specific shoppers are, what their preferences may be, and how they’re interacting within the store environment, they can fine-tune messaging to offer truly relevant promotions.

In essence, in-store retail media turns screens into strategic points of engagement, allowing brands to connect with the consumer in a far more relevant and powerful manner than what is possible with DOOH.

Driving the path to purchase

In-store retail media does more than simply display ads - it actively influences the purchasing decision.

While a static digital sign may attempt to raise brand awareness or reinforce marketing campaigns, in-store retail media offers a fully contextualised journey for shoppers - delivering attributable results and the ability to get analytics from the aisles to better inform future campaigns and prove incremental sales.

By strategically placing smart digital screens at key points within the store, such as near product displays or at checkout, brands can deliver relevant messages that complement the shopper journey and have stronger influence.

For example, a customer may be on the fence about buying a particular product, and a nearby screen could highlight a limited time offer or provide additional information about the product’s benefits.

This kind of last minute influence, right at the Point of Sale, transforms passive ad displays into active sales drivers. In fact, in-store retail media is more akin to dynamic merchandising, offering a measurable boost to sales, which traditional DOOH cannot achieve in isolation.

Integration with the digital ecosystem

In-store retail media bridges the gap between physical retail environments and online marketing strategies. Screens in stores are no longer just outlets for broadcasting ads—they are now integrated touchpoints within a broader omnichannel strategy.

With technologies like data-compliant sensors and real-time data analytics, in-store retail media can garner and use a wealth of in-store customer behavioural data. This data can then be fed back into digital marketing campaigns, enabling retailers to optimise both their in-store and online efforts.

For instance, a customer who interacts with a product in-store but does not purchase it could be retargeted with personalised ads online, perhaps through email, social media,mobile app notifications or connected television.

Moreover, in-store retail media can be tied into loyalty programmes and CRM systems, allowing retailers to deliver highly specific, data-driven content that enhances the shopping experience, creating convenience and minimising frustrations for their customers. 

Advanced metrics and real-time feedback

One of the main limitations of DOOH is the difficulty in tracking its effectiveness. While impressions can be roughly estimated based on foot traffic or location, there is little concrete data on how many people actually engage with the ad or how it influences behaviour. In-store retail media, however, offers the possibility of real-time feedback and advanced metrics.

With technologies like RFID, beacons, and customer analytics platforms, retailers can track exactly how many people interact with specific in-store media placements, which products they pick up afterwards, and whether the digital content influenced their purchasing decision. This closed-loop attribution provides a clear understanding of ROI and allows for real-time optimisation of campaigns.

For instance, if a particular promotion isn’t generating the expected engagement, the messaging can be adjusted on the fly to better suit the audience. This ability to measure, analyse, and act on consumer behaviour swiftly sets in-store retail media apart from the more static nature of traditional DOOH.

Creating an immersive experience

DOOH typically plays a supporting role in the marketing mix, adding visual presence but rarely creating a fully immersive experience. In-store retail media, by contrast, has the ability to enhance the overall shopping environment, turning it into an interactive and engaging experience.

By blending digital signage with interactive displays, touchscreens, and augmented reality, in-store retail media can provide valuable information and entertainment that enriches the customer journey.

Consider a shopper in a clothing store. Instead of simply seeing a screen that shows a rotating set of ads, they might use an interactive display to browse additional product options, check sizes, or even virtually try on items using AR technology.

This level of interactivity not only makes the shopping experience more enjoyable but also encourages customers to spend more time in the store and engage with the brand in a more meaningful way. For the retailer and the brands they supply, these interaction provide useful information on how to serve shoppers better in the future.

Successful retail media: beyond just screens

In-store retail media is so much more than just another form of digital out-of-home advertising. It represents a powerful, data-driven tool for personalising customer experiences, influencing purchasing decisions, and integrating with the broader digital marketing ecosystem.

While DOOH may still play a valuable role in reaching broader audiences, in-store retail media takes things to the next level by creating dynamic, interactive, and measurable touchpoints that can directly impact sales.

As retailers continue to invest in digital transformation, the role of in-store retail media will only become more central to their strategy. Its ability to seamlessly blend the physical and digital worlds, deliver personalised content, and provide advanced metrics makes it a cornerstone of the future of retail.

It’s time to stop thinking of it as just another screen and start recognising it as a critical component of a modern, integrated marketing strategy.

09 Nov 16:32

Avocats et juristes préoccupés par la GenAI

Sans surprise, 3 des 5 tendances qui devraient influencer le secteur juridique dans les 3 prochaines années sont liées à l'IT : l'IA (...)
09 Nov 16:24

Zalando améliore sa cabine d'essayage virtuelle avec un avatar 3D

Pour la première fois, Zalando intègre sa technologie de mesure corporelle à sa cabine d'essayage en ligne, offrant une expérience toujours plus personnalisée. Ce projet permet d'essayer virtuellement une sélection de produits Levi's pour hommes et femmes tout en limitant le nombre de retours produits.
09 Nov 16:19

OpenAI révèle comment utiliser ChatGPT-4o plus efficacement

by Yohann Poiron

OpenAI vient de révéler une approche complète pour tirer le meilleur parti de son modèle ChatGPT-4o, qui se distingue par sa polyvalence et sa capacité à simplifier des workflows pour une multitude de tâches professionnelles. En collaboration avec le guide de TheAIGRID, conçu pour optimiser l’utilisation de ChatGPT-4o, les entreprises de toutes tailles peuvent désormais […]

L’article OpenAI révèle comment utiliser ChatGPT-4o plus efficacement est apparu en premier sur BlogNT : le Blog des Nouvelles Technologies.

09 Nov 16:15

TSMC will stop making 7 nm chips for Chinese customers

by Kathrin Hille and Ryan McMorrow, Financial Times

Taiwan Semiconductor Manufacturing Company has notified Chinese chip design companies that it will suspend production of their most advanced artificial intelligence chips, as Washington continues to impede Beijing’s AI ambitions.

TSMC, the world’s largest contract chipmaker, told Chinese customers it would no longer manufacture AI chips at advanced process nodes of 7 nanometers or smaller as of this coming Monday, three people familiar with the matter said.

Two of the people said any future supplies of such semiconductors by TSMC to Chinese customers would be subject to an approval process likely to involve Washington.

Read full article

Comments

09 Nov 16:11

Google rolls out its Gemini AI-powered video presentation app

by Umar Shakir
Vids help me create screen with a prompt that says “create a market research overview on Gen Z retail trends using [gen Z research report doc]”
Google Vids’ “Help me create” screen powered by Gemini AI. | Image: Google

Google is generally rolling out its Gemini AI-powered Vids app that lets you create video presentations using a prompt. Some of Vids’ key features include letting Gemini auto-insert stock footage for you, generating a script, and making AI voiceovers so you don’t have to speak. Google advertises that the tool can help turn customer support articles into videos, make training videos, share company announcements, create meeting recaps, and more.

Vids will be available by default for Workspace organizations with access, but Google notes possible usage limits may apply to features like “Help me create” and AI voiceovers starting in 2026.

First announced in April, Vids lets you drop in relevant docs, slides, voiceovers, and video recordings...

Continue reading…

09 Nov 16:10

Quantum encoding methods could slash circuit complexity in machine learning

A recent study by researchers from CSIRO and the University of Melbourne has made progress in quantum machine learning, a field aimed at achieving quantum advantage to outperform classical machine learning.
09 Nov 16:09

Artificial magnetic muscles can support tensile stresses up to 1,000 times their own weight

A research team, led by Professor Hoon Eui Jeong from the Department of Mechanical Engineering at UNIST has introduced an innovative magnetic composite artificial muscle, showcasing an impressive ability to withstand loads comparable to those of automobiles. This material achieves a stiffness enhancement of more than 2,700 times compared to conventional systems. The study is published in Nature Communications.
09 Nov 16:07

Claude AI to process secret government data through new Palantir deal

by Benj Edwards

Anthropic has announced a partnership with Palantir and Amazon Web Services to bring its Claude AI models to unspecified US intelligence and defense agencies. Claude, a family of AI language models similar to those that power ChatGPT, will work within Palantir's platform using AWS hosting to process and analyze data. But some critics have called out the deal as contradictory to Anthropic's widely-publicized "AI safety" aims.

On X, former Google co-head of AI ethics Timnit Gebru wrote of Anthropic's new deal with Palantir, "Look at how they care so much about 'existential risks to humanity.'"

The partnership makes Claude available within Palantir's Impact Level 6 environment (IL6), a defense-accredited system that handles data critical to national security up to the "secret" classification level. This move follows a broader trend of AI companies seeking defense contracts, with Meta offering its Llama models to defense partners and OpenAI pursuing closer ties with the Defense Department.

Read full article

Comments

06 Nov 23:51

On ne devrait pas être promu manager si on ne sait pas utiliser ChatGPT

by Bertrand DUPERRIN

Un manager donne une instruction, confie une mission, et au moment de présenter son travail ou débrieffer de ses résultats le collaborateur s’avance assez confiant, satisfait, voire plus du travail qu’il a réalisé.

Et là catastrophe ! Rien ne va, ça ne correspond pas aux attentes et il se fait vertement sermonner. Une chose d’autant plus désagréable qu’il ne s’y attend pas du tout, voire comptait bien être félicité.

C’est une situation que nous avons tous soit observé soit vécu personnellement soit les deux. Très désagréable quand vous êtes directement concerné, gênant quand vous en êtes témoin.

Manager trop exigeant ? Salarié incompétent ?

Ca serait un raccourci un peu facile car le plus souvent ils ne se sont juste pas compris.

L’idée ici est de réfléchir à comment être certains d’être compris des autres quand on formule une demande et vous verrez que ChatGPT peut être sinon un bon professeur pour vous en tout cas un bon cobaye pour tester votre capacité à communiquer de manière claire et compréhensible.

La communication manager/managé est très perfectible

Quelques exemples.

Le mémo n’était pas assez fouillé ni bien présenté mais il n’a jamais été dit qu’il devait servir en board et que le manager avait une idée particulière en tête.

Le document est arrivé en retard mais le collaborateur ne savait pas que le directeur n’attendait que quelques bullet points pour se faire une idée sur un sujet à défricher et que cette présentation n’avait aucunement vocation à être diffusée.

En réglant un conflit avec un client, un fournisseur ou même un conflit interne le collaborateur n’a fait que mettre de l’huile sur le feu ou a mal protégé les intérêts de l’entreprise ou les a tellement bien protégés qu’ils ont perdu le client. Mais on ne lui a jamais expliqué sa marge de négociation ni l’issue attendue à part « régler le problème ».

Le livrable est de bonne qualité à première vue mais pas assez fouillé. D’un autre côté on avait dit au collaborateur « fais moi un topo là dessus ».

Le collaborateur ne connait pas tous les chiffres d’un projet dans les détail avec les taux de rentabilité et d’utilisation de chaque ressource etc, juste les données économiques macro et se fait passer un savon. D’un autre côté on lui avait dit « faisons un point rapide« .

L’office manager organise le pot de rentrée mais c’est soit pas assez qualitatif soit il n’y en a pas assez soit c’est trop cher. Mais on lui avait dit « organise le pot ».

La même office manager organise le déplacement d’un dirigeant. Train trop tôt, retour à une heure où il a une réunion, l’hôtel ne plait pas. Elle passe un sale quart d’heure mais on lui a juste dit d’organiser le déplacement.

Des exemples comme cela on en a tous des dizaines voire plus en mémoire.

Ce qui se passe en fait c’est que le manager ou le dirigeant sait exactement ce qu’il attend, il se dit à lui même « il faut faire ça » et ensuite délègue à quelqu’un en formulant la chose de la même manière.

Vous n’avez jamais remarqué ? Que ce soit en cours ou en fin de réunion ou dans un email ou un tchat lapidaire (le célèbre « merci de …. ») la tâche/mission est confiée en une phrase, une ligne sans autre indication de contexte.

Normal : le manager sait ce qu’il veut mais oublie que la personne à qui il délégue n’est pas lui.

Normal mais totalement inefficace car la personne se retrouve face à deux choix :

1°) Essayer de deviner tout ce qui n’a pas été dit

2°) Demander des précisions au risque de laisser penser qu’elle n’a pas compris, laisser penser qu’elle n’est pas compétente ou, pire, de se faire rabrouer parce que le manager « n’a pas le temps ».

Le manager Vs. ChatGPT

J’entends beaucoup de managers se réjouir de l’arrivée de l’IA car elle va leur permettre d’avoir rapidement les informations et analyses qu’ils n’arrivent pas à avoir de salariés qui ne comprennent jamais rien et sont des crétins.

S’ils ne l’ont pas fait je leur conseille d’essayer dès maintenant.

Quiconque a commencer à se frotter à l’IA, même pour des choses simples, a vite compris quelque chose : si la demande n’est pas claire on obtient quelque chose de qualité discutable et il faut faire de multiples repasses pour affiner et arriver à quelque chose d’acceptable.

J’ai lu une analogie très intéressante sur les IA il y a peu :

« Lorsque j’ai commencé à jouer avec ChatGPT, je n’étais pas sûr de l’usage que j’en ferais. Puis, un jour, alors que je travaillais à distance, j’ai commencé à taper des questions, m’adressant comme je le ferais à un collègue assis en face de moi et m’aidant à réfléchir à de nouvelles idées. Il en est résulté une conversation passionnante qui m’a permis de réduire mon temps de travail et d’améliorer mon produit final. Après m’être familiarisé avec l’outil, j’ai commencé à le considérer comme un partenaire de réflexion omniprésent, ce qui m’a amené à réfléchir à la manière d’intégrer la GenAI dans des équipes non techniques. J’ai réalisé que l’intégration de GenAI n’est pas si différente de l’ajout d’un nouveau membre à l’équipe. »

Fast Company – How to transition nontechnical teams to use GenAI

J’ai vraiment trouvé l’idée de faire comme si on ajoutait un nouveau membre à l’équipe très juste et pertinente.

Il y aussi l’idée l’implicite qu’il a compris qu’il devait se faire comprendre de l’IA.

Et bien votre collaborateur ou votre stagiaire n’est pas si différent de ChatGPT : si vous ne vous faites pas comprendre de lui il ne pourra rien faire de bien pour vous.

Mais pour beaucoup de manager une différence existe dans leur tête : ils ont compris qu’il fallait dompter ChatGPT alors qu’ils partent du principe que leurs collaborateurs vont naturellement comprendre sans connaitre les tenants et les aboutissants

On pourrait même dire que l’absence de lien hiérarchique avec l’IA crée une relation plus équilibrée pour ne pas dire saine.

Et pour en revenir au fait que certains managers pensent que tous leurs collaborateurs sont mauvais on a plutôt la preuve que leur incompétence présumée est proportionnelle à l’incapacité du manager à se faire comprendre.

Sachez prompter vos équipes

Donc le manager qui a compris que c’était à lui de faire l’effort d’être compris par ChatGPT a appris à lui parler, à lui donner des instructions. Ce qui dans le jardin de l’IA se dit écrire un prompt ou prompter.

Sans avoir la prétention d’être un expert sur le sujet qui est beaucoup moins simple qu’il ne semble à prime abord, voici quelques règles pour un bon prompt :

Clarté et spécificité : Plus le prompt est spécifique, plus la réponse sera pertinente. Il est important de bien préciser le contexte et les attentes. Par exemple, au lieu de demander « Explique l’IA », on pourrait dire « Explique comment l’IA améliore la productivité dans les entreprises ».

Contexte : Inclure des informations contextuelles qui peuvent aider à affiner la réponse. Par exemple, « Décris les avantages de l’IA dans le secteur de la santé en 2024 » offre un cadre plus précis. Exemple « c’est pour une présentation à mon patron » ou « c’est pour écrire un article de vulgarisation sur linkedn »

Objectif : Dire à quoi va servir la réponse aide à la rendre plus pertinente sur le fonds et la forme. Exemple « c’est pour une présentation à mon patron » ou « c’est pour écrire un article de vulgarisation sur linkedn »

Format souhaité : Indiquer le type de réponse attendu (paragraphe, liste à puces, tableau, etc.) ainsi que la longueur permet d’orienter la réponse. Par exemple, « Fais une liste des avantages et inconvénients du télétravail ».

Désambiguïsation : Éviter les formulations vagues ou ambiguës pour réduire le risque de recevoir une réponse hors sujet.

Tonalité et style : Préciser la tonalité (formelle, informelle, technique) ou le style (persuasif, descriptif) si c’est pertinent pour le contexte.

Je pense que d’autres pourront fournir un guide technique très précis de l’écriture du prompt parfait mais ma conviction profonde est que si au lieu de donner une instruction lapidaire d’une ligne le manager prenait le temps de donner les éléments de contexte, préciser ses attentes sur le fonds et la forme, ses intentions et d’éventuelles contraintes il serait moins déçu par le travail de ses collaborateurs.

L’auteur de l’article disait qu’il fallait considérer l’IA comme un nouveau membre de l’équipe mais d’une certaine manière il faudrait peut être considérer certains collaborateurs comme une IA, surtout les moins expérimentés et ceux avec qui on a le moins de vécu commun.

Mais peu importe le référentiel dont on se sert l’idée est bien que, entre être humains, on surestime souvent soit la capacité des autres à deviner nos pensées ou notre propre capacité à être clairs.

Leçons tirées de l’agilité

je faisais référence à ChatGPT car c’est un sujet à la mode mais par le passé j’avais été inspiré par quelque chose de plus ancien lorsqu’il s’agissait de formaliser des instructions : les user stories.

Leur principe repose sur la création de courtes descriptions ou scénarios exprimant les besoins d’un utilisateur ou d’un client dans un projet, souvent dans le contexte du développement logiciel, en se concentrant sur la valeur ajoutée qu’une fonctionnalité ou un service doit apporter.

Si c’est dans l’industrie du développement que j’y ai été exposé et sensibilisé, j’ai vite compris qu’on pouvait les appliquer à autre chose.

Lors de la mise en place d’une démarche d’amélioration continue qui n’avait rien à voir avec un quelconque développement j’avais identifié l’agilité comme critère d’acceptabilité de la démarche avec ceux qu’elle allait concerner : avancer vite par petites itérations en créant ensemble la vision cible.

Mais au moment de distribuer les tâches à chacun (moi y compris) il importait qu’on soit vraiment alignés sur les attentes. Toutes les personnes concernées pratiquant l’agilité dans le cadre de leurs fonctions on a logiquement fini avec une formulation proche de celle d’une user story, quand bien même l’objectif n’était pas de développer quoi que ce soit mais davantage de modifier des process et des choses liées à l’organisation et méthodes de travail.

Je rappelle les caractéristiques d’une user story pour ceux à qui le concept est étranger.

Perspectives centrées sur l’utilisateur : Elles sont écrites du point de vue de l’utilisateur final (client, employé, etc.) afin de s’assurer que le résultat final répond à ses besoins réels.

Simples et concises : Les user stories sont intentionnellement courtes et claires, sans entrer dans les détails techniques. Elles décrivent ce que l’utilisateur souhaite accomplir et pourquoi cela est important pour lui.

Centrée sur la valeur métier : L’accent est mis sur la valeur que la fonctionnalité ou l’amélioration apportera à l’utilisateur ou à l’organisation.

Itératives et évolutives : Les user stories sont souvent complétées ou ajustées au fil du projet. Elles évoluent à mesure que l’équipe développe une meilleure compréhension des besoins et des priorités de l’utilisateur.

Critères d’acceptation : Chaque user story est généralement accompagnée de critères d’acceptation, qui sont des conditions spécifiques qui doivent être remplies pour que l’histoire soit considérée comme complétée.

Cela peut sembler très formel mais ça a le mérite d’être exhaustif et si vous formulez une demande de cette manière le risque que ce que vous obteniez soit aux antipodes de vos attentes est faible, peu importe le domaine auquel vous l’appliquez.

Mais autant on sent intuitivement que c’est à nous de faire les efforts pour être compris par une IA, que dans le cadre d’un projet aux multiples intervenants provenant de différents métiers il faut une manière très simple mais structurée de formuler ses attentes, autant lorsqu’on s’adresse à une autre personne on a tendance à lui passer la patate chaude sans faire d’effort et, sans surprise, on est souvent déçu par le résultat.

Conclusion

Par manque de temps, de respect, d’attention ou je ne sais quelle raison trop de managers ont tendance à être expéditifs dans leurs demandes quitte à être incompris, être déçu par le résultat et finalement reporter la faute sur les autres.

Ou peut être ne savent ils pas ce qu’ils veulent et espèrent que quelqu’un trouvera pour eux ?

Quoi qu’il en soit un effort doit être fait pour qu’instructions et demandes soient formulées de manière claire et non ambiguë.

Je ne dis pas que si on est capable d’écrire un bon prompt on sera un bon manager mais j’ai des doutes que quelqu’un qui en soit incapable soit capable de se faire comprendre et donner des instructions claires.

Image : manager utilisant chatGPT de Celia Ong via Shutterstock

L’article On ne devrait pas être promu manager si on ne sait pas utiliser ChatGPT est apparu en premier sur Bloc-Notes de Bertrand Duperrin.

06 Nov 23:47

OpenAI acquired Chat.com

by Kyle Wiggers

OpenAI bought Chat.com, adding to its collection of high-profile domain names. As of this morning, Chat.com now redirects to OpenAI’s AI-powered chatbot, ChatGPT. An OpenAI spokesperson confirmed the acquisition via email. Chat.com is one of the older domains on the web, having been registered in September 1996. Last year, it was reported that HubSpot co-founder […]

© 2024 TechCrunch. All rights reserved. For personal use only.

06 Nov 23:46

The best note-taking apps for collecting your thoughts and data

by Barbara Krasnoff
Evernote
Photo by James Bareham / The Verge

These days, we are all dealing with huge amounts of information, from meeting notes to social media, to photos and videos, to whatever else we’ve collected — and we are all trying to find some way to store it, organize it, and find it when we need it.

If you want to get really basic, you can use a spreadsheet or create a simple set of word-processing documents. Otherwise, you can try what is somewhat inaccurately described as a note-taking app. These apps, at their simplest, store your notes and other thoughts and, at their more complex, are capable of manipulating any and all content you want to drop into them.

One of the most well-known has been Evernote, which has gone through quite a few changes over the years, both financially and...

Continue reading…

06 Nov 23:39

Nearly three years since launch, Webb is a hit among astronomers

by Stephen Clark

From its halo-like orbit nearly a million miles from Earth, the James Webb Space Telescope is seeing farther than human eyes have ever seen.

In May, astronomers announced that Webb detected the most distant galaxy found so far, a fuzzy blob of red light that we see as it existed just 290 million years after the Big Bang. Light from this galaxy, several hundreds of millions of times the mass of the Sun, traveled more than 13 billion years until photons fell onto Webb's gold-coated mirror.

A few months later, in July, scientists released an image Webb captured of a planet circling a star slightly cooler than the Sun nearly 12 light-years from Earth. The alien world is several times the mass of Jupiter and the closest exoplanet to ever be directly imaged. One of Webb's science instruments has a coronagraph to blot out bright starlight, allowing the telescope to resolve the faint signature of a nearby planet and use spectroscopy to measure its chemical composition.

Read full article

Comments

05 Nov 22:11

Repurposed RFID tags allow for battery-free sensing and tracking

Data is power. According to Dinesh Bharadia, an associate professor at UC San Diego in the Department of Electrical and Computer Engineering with an affiliate appointment in the Department of Computer Science and Engineering and the Qualcomm Institute (QI), "data will be the next decade's 'silicon.'"
05 Nov 21:45

Alan recrute un assistant médical virtuel

by Patrice
Alan
Depuis ses origines, Alan appréhende la santé non pas seulement par le volet de l'assurance, qui constitue son métier fondamental, mais comme une expérience globale. Afin d'améliorer l'accompagnement médical offert par ses professionnels affiliés, elle leur adjoint désormais une sorte d'assistant stagiaire piloté par l'intelligence artificielle.

Parmi les nouveautés dévoilées par la jeune pousse hexagonale pour cet automne 2024 figurent, dans le registre de la prévention, une boutique d'une centaine de produits soigneusement sélectionnés, français et vendus à prix négociés, contribuant au maintien de la forme physique – depuis les compléments alimentaires jusqu'aux appareils connectés – et un programme de comptage des pas assorti de quelques fonctions ludiques destinées à encourager la marche à pied. Et puis il y a Mo.

Mo est un interlocuteur virtuel qui s'invite dans l'espace de tchat qu'Alan a déployé dans son application dans le but de permettre à ses assurés d'interroger des médecins, généralistes ou spécialistes, sur leurs petites inquiétudes du quotidien (il ne s'agit pas de consultations à proprement parler). Il devient donc le premier contact au démarrage d'un échange et propose de répondre instantanément aux questions qui lui sont posées, tout en laissant en permanence la possibilité de transférer la conversation à un humain.

L'avantage pour les utilisateurs est évident : au lieu d'attendre la disponibilité d'un interlocuteur, promise sous 15 minutes en général (mais quelquefois plus dans les faits), l'agent intelligent offre un retour immédiat, qui réduit bien des frustrations. Du point de vue des praticiens, la promesse de n'avoir plus à prendre en charge que des problématiques un peu plus complexes, qui n'entrent pas dans le champ de compétences de la machine, constitue également un motif de satisfaction.

Alan – Mo

Cependant, bien sûr, l'introduction d'un moteur d'IA prodiguant des conseils médicaux a de quoi susciter des inquiétudes quand on connaît la propension des algorithmes, quelle que soit leur qualité, à produire des hallucinations et autres anomalies. Afin de rassurer les patients autant que pour éviter des erreurs, Alan implémente un contrôle systématique : toutes les « prescriptions » automatiques sont relues par un médecin et sont soit confirmées, soit corrigées sous un quart d'heure. On peut supposer que ces vérifications servent aussi à entraîner en continu le modèle mis en œuvre.

Bien qu'on puisse soupçonner un biais d'échantillon, les résultats obtenus lors de la phase de test organisée sur quelques mois semblent montrer une confiance surprenante de la part des assurés, puisque 4 sur 5 acceptent d'interagir avec Mo plutôt qu'avec l'humain qu'ils souhaitaient consulter initialement et ils apprécient largement l'expérience, en lui attribuant une note moyenne de 4,6 sur 5. La validation par les professionnels de 93% des réponses fournies n'y est probablement pas étrangère.

Avec cette initiative, qui n'en est qu'à ses prémices, Alan démultiplie sa couverture des besoins de ses clients en matière de santé et de bien-être, en complément de son rôle primaire de remboursement des soins. La démarche est nécessairement prudente mais, avec la maturité, elle devrait parvenir à créer de nouveaux comportements dont toutes les parties prenantes tireront les bénéfices, les uns pour leur qualité de vie, les autres (employeurs) en termes de productivité et elle-même sur sa rentabilité.
05 Nov 21:09

China's New Spaceship Is a Blatant Ripoff of SpaceX's Starship

by Victor Tangermann
China has showed off a concept for its upcoming Long March 9, a ultra heavy-lift reusable rocket that looks a lot like SpaceX's Starship.

AliExpress Starship

China has showed off a concept for its upcoming Long March 9, a ultra heavy-lift reusable rocket designed to deliver the first Chinese astronauts to the surface of the Moon in the 2030s.

And as Ars Technica reports, its design is an unabashed ripoff of SpaceX's Starship rocket, also currently in development. A slide from a presentation at a recent airshow in China shows off a strikingly familiar design for a "two-stage, fully reusable configuration" — right down to aerodynamic flaps that aid the rocket during its descent back to the surface.

The rocket will purportedly feature 30 methane and liquid oxygen fuel rocket engines, not unlike the 33 Raptor engines that power Starship's first stage and which rely on the same mix of fuel.

While the country isn't planning to fly the massive, 374-feet-tall rocket until 2033, the design highlights just how much of a headstart and influence SpaceX has in the current race to develop reusable rockets.

Sincerest Flattery

As Ars points out, it's far from the first time we've seen China's space program — as well as Chinese space startups — rip off SpaceX's designs. In 2021, a promotional video by the China Academy of Launch Vehicle Technology clearly aped the Elon Musk-led company's early plans for an end-to-end transportation system that made use of Starship, which was known as "BFR" at the time.

Another suspicious Starship lookalike design for a two-stage methane-liquid launch system surfaced in 2022 as well.

Just last week, Chinese space launch company Cosmoleap released a promo video that copied SpaceX's plans for a tower with "chopstick" arms designed to catch the first stage of a reusable rocket, as spotted by SpaceNews. SpaceX recently pulled off the feat for the first time during its fifth Starship test flight last month.

A different Chinese space startup, called Space Pioneer, also announced plans for a rocket late last year that closely resembles SpaceX's reusable Falcon 9.

In short, is this all just a sign that China has run out of ideas and is blindly following its international competition? Or is it a case of "if it ain't broke, don't fix it"?

Regardless, it's a strategy. China is still hoping to land astronauts on the Moon by 2030 using its Long March 10 rocket, a predecessor to its Starship-like Long March 9 rocket.

That's several years after NASA's first scheduled crewed landing, which is tentatively scheduled for late 2026.

But to establish a more permanent presence on the Moon, a reusable rocket would give either country a massive leg up, whether it's a genuine Starship or a lookalike.

More on Starship: In Video of Elon Musk Playing Diablo IV, Background Gives Away Wild Details About Starship

The post China's New Spaceship Is a Blatant Ripoff of SpaceX's Starship appeared first on Futurism.