Jean-Philippe Encausse
Shared posts
Ce matériau inspiré du caméléon pourrait rendre l'invisibilité réelle 🦎
LG’s transparent OLED TV will put a very visible $60,000 hole in your wallet
LG’s Signature OLED T, the transparent TV that was the company’s big flashy reveal at CES 2024, now has an official price: $59,999. It’s available beginning today in the United States “at select LG-authorized retailers including Best Buy, Video & Audio Center and via LG.com” the company said in a press release. You can’t just walk into your local Best Buy and get this thing, mind you; it’s a special order kind of situation. LG says the first purchase will be made later tonight at a gala at Video & Audio Center.
The OLED T lets you switch between its mind-boggling transparent mode, where you can clearly see whatever’s behind the OLED panel, and an opaque mode that eliminates distractions when enjoying content on the 77-inch screen.
As is usually the case with Signature models, this OLED gets some completely unique software tricks that you won’t find on other LG sets. These include:
T-Objet: an Always-On-Display (AOD) mode that transforms the screen into a transparent digital canvas, ideal for showcasing artwork, videos or photos in outstanding color and clarity.
T-Bar: a sleek info-ticker that appears along the bottom edge of the screen, providing sports results, IoT device statuses, weather forecasts or song title information. The rest of the screen remains unused when T-Bar is active, presenting a clean, uncluttered look and a clear view of the space behind LG’s “virtually invisible” display. For added convenience,
T-Home: offers a user-friendly interface that delivers a well-organized overview of available services, as well as quick access to apps, settings and other features.
The OLED T is forever out of reach for mere peasants like myself, but even if I had $60,000 burning a hole in my pocket, I’m not sure the transparency trick would be enough for me to hand it over. The far more conventional LG G4 would do just fine. But if you lead a lavish lifestyle and have the perfect scenario dreamed up for this thing, it’s now actually available to buy.
Measuring a Well with Just a Hammer and a Smartphone
What’s the best way to measure the depth of a well using a smartphone? If you’re fed up with social media, you might kill two birds with one stone and drop the thing down the well and listen for the splash. But if you’re looking for a less intrusive — not to mention less expensive — method, you could also use your phone to get the depth acoustically.
This is a quick hack that [Practical Engineering Solutions] came up with to measure the distance to the surface of the water in a residential well, which we were skeptical would work with any precision due to its deceptive simplicity. All you need to do is start a sound recorder app and place the phone on the well cover. A few taps on the casing of the well with a hammer send sound impulses down the well; the reflections from the water show up in the recording, which can be analyzed in Audacity or some similar sound editing program. From there it’s easy to measure how long it took for the echo to return and calculate the distance to the water. In the video below, he was able to get within 3% of the physically measured depth — pretty impressive.
Of course, a few caveats apply. It’s important to use a dead-blow hammer to avoid ringing the steel well casing, which would muddle the return signal. You also might want to physically couple the phone to the well cap so it doesn’t bounce around too much; in the video it’s suggested a few bags filled with sand as ballast could be used to keep the phone in place. You also might get unwanted reflections from down-hole equipment such as the drop pipe or wires leading to the submersible pump.
Sources of error aside, this is a clever idea for a quick measurement that has the benefit of not needing to open the well. It’s also another clever use of Audacity to use sound to see the world around us in a different way.
New physics sim trains robots 430,000 times faster than reality
On Thursday, a large group of university and private industry researchers unveiled Genesis, a new open source computer simulation system that lets robots practice tasks in simulated reality 430,000 times faster than in the real world. Researchers can also use an AI agent to generate 3D physics simulations from text prompts.
The accelerated simulation means a neural network for piloting robots can spend the virtual equivalent of decades learning to pick up objects, walk, or manipulate tools during just hours of real computer time.
"One hour of compute time gives a robot 10 years of training experience. That's how Neo was able to learn martial arts in a blink of an eye in the Matrix Dojo," wrote Genesis paper co-author Jim Fan on X, who says he played a "minor part" in the research. Fan has previously worked on several robotics simulation projects for Nvidia.
Home Assistant’s Voice Preview Edition is a little box with big privacy powers
Home Assistant announced today the availability of the Voice Preview Edition, its own design of a living-room-friendly box to offer voice assistance with home automation. Having used it for a few weeks, it seems like a great start, at least for those comfortable with digging into the settings. That's why Home Assistant is calling it a "Preview Edition."
Using its privacy-minded Nabu Casa cloud—or your own capable computer—to handle the processing, the Voice Preview Edition (VPE) ($60/60 euros, available today) has the rough footprint of a modern Apple TV but is thinner. It works similarly to an Amazon Echo, Google Assistant, or Apple Siri device, but with a more focused goal. Start with a wake word—the default, and most well-trained version, is "Okay, Nabu," but "Hey, Jarvis" and "Hey, Mycroft" are available. Follow that with a command, typically something that targets a smart home device: "Turn on living room lights," "Set thermostat to 68," "Activate TV time." And then, that thing usually happens.
"That thing" is primarily controlling devices, scenes, and automations around your home, set up in Home Assistant. That means you have to have assigned them a name or alias that you can remember. Coming up with naming schemes is something you end up doing in big-tech smart home systems, too, but it's a bit more important with the VPE.
The Three Bricklayers Story
The Three Bricklayers’ story illustrates the power of purpose.
What is the 3 Bricklayers Story?
A simple version goes that a person walked past a building project and asked three workers the same question: “What are you doing?”
The first replied, “I’m laying bricks.”
The second replied, “I’m building a wall.”
And the third replied, “I’m creating a cathedral.”
The story highlights how we can view our work differently depending on whether we focus on the immediate task, the short-term goal, or the larger vision. The first worker focuses on the task at hand, the second sees the outcome of their work, and the third connects to the broader purpose of the project.
Various tellings have the first worker hunched over or working slowly. The second and third workers take increasing pride in their work, often achieving more.
Finding Balance in Work
There’s value in all three perspectives. There can be a lot of pride and skill in laying bricks—or whatever your equivalent task is—as well as it can be done. Setting clear, intermediate goals keeps progress on track. And someone who spends all their time looking at plans or daydreaming about what the building will become may not lay bricks as well as they need to.
To do something well, we probably need a balance of all three aspects:
- Pride and skill in detail and craft
- Progress through clear intermediate goals
- Vision and meaning for our work
Understanding what I’m working towards and believing it’s worthwhile is a powerful motivator for me when the going gets tough.
This post isn’t really about cathedrals, but I studied the brilliant Catalan architect Antoni Gaudí in my teens. As a real-life cathedral metaphor, his incredible Basílica de la Sagrada Família in Barcelona is a striking example. Gaudí’s vision for the basilica has inspired generations of artisans and workers since he took over the project in 1882. Execution, however, has been a challenge, with construction ongoing today. But that hasn’t stopped it from inspiring and drawing in visitors for decades.
I like the three bricklayers parable as a reminder that when I’m grinding on something, it helps to reconnect with the why behind my effort.
Origins of the 3 Bricklayers Parable
Like many parables, this story has been told in different forms. An early version appears in Bruce Barton’s 1927 book What Can a Man Believe (p252), featuring Sir Christopher Wren, the architect of St Paul’s Cathedral in London after the Great Fire of 1666.
“One morning he passed among the workmen, most of whom did not know him, and of three different men engaged in the same kind of work he asked the same question: ‘What are you doing?’
From the first he received the answer: ‘I am cutting this stone.’
From the second the answer was: ‘I am earning three shillings and six pence a day.’
But the third man straightened up, squared his shoulders, and holding his mallet in one hand and chisel in the other, proudly replied: ‘I am helping Sir Christopher Wren to build this great cathedral.’”
It’s hard to believe it happened, but that doesn’t mean it can’t be helpful.
Related Ideas to the 3 Bricklayers Story
Also see:
2024 Recap
We’re reaching the end of the year, so here’s our recap of 2024. Unsurprisingly, several of the projects we explored this year had something to do with Artificial Intelligence, whether it’s innovative product developments or fictional interpretations of AI.
There’s no doubt that AI is going to be a part of our world moving forward so we will likely be seeing more explorations of this in future posts.
A special heartfelt thanks to everyone who reached out, contributed, bought a book, bought some merch or donated. All your kindness and friendliness has kept us going. Thanks for being cool people!
Have a nice holiday break and we’ll catch you next year for more FUI and future concepts in 2025!
Ulta Beauty taps NVIDIA StyleGAN2 gen AI model for virtual hair style and colour try-on experience
A new Ulta Beauty AI app uses selfies to show near instant, realistic previews of desired hairstyles.
GLAMlab Hair Try On lets users take a photo, upload a headshot or use a model’s picture to experiment with different hair colours and styles. The experience is powered by the NVIDIA StyleGAN2 generative AI model.
The hair colour try-ons feature links to Ulta Beauty products so shoppers can achieve the look in real life. Ulta Beauty, which has more than 1,400 stores across the US, has found that people who use the virtual tool are more likely to purchase a product than those who don’t.
“Shoppers need to try out hair and makeup styles before they purchase,” says Juan Cardelino, Director of the Computer Vision and Digital Innovation Department at Ulta Beauty. “As one of the first cosmetics companies to integrate makeup testers in stores, offering try-ons is part of Ulta Beauty’s DNA - whether in physical or digital retail environments.”
GLAMlab is Ulta Beauty’s first generative AI application, developed by its digital innovation team.
To build its AI pipeline, the team turned to StyleGAN2, a style-based neural network architecture for generative adversarial networks, aka GANs. StyleGAN2, developed by NVIDIA Research, uses transfer learning to generate infinite images in a variety of styles.
“StyleGAN2 is one of the most well regarded models in the tech community, and, since the source code was available for experimentation, it was the right choice for our application,” Cardelino says. “For our hairstyle try-on use case, we had to license the model for commercial use, retrain it and put guardrails around it to ensure the AI was only modifying pixels related to hair - not distorting any feature of the user’s face.”
Available on the Ulta Beauty website and mobile app, the hair style and colour try-ons rely on NVIDIA Tesnor Core GPUs in the cloud to run AI inference, which takes around five seconds to compute the first style and about a second each for subsequent styles.
The company next plans to incorporate virtual trials for additional hair categories like wigs and is exploring how the virtual hairstyle try-ons could be connected to in-store styling services.
“Stylists could use the tool to show our guests how certain hairstyles will look on them, giving them more confidence to try new looks,” Cardelino says. “Hair and makeup are playful categories. Virtual try-ons are a way to explore options that may be out of a customer’s comfort zone without needing to commit to a physical change.”
RTIH AI in Retail Awards
RTIH, organiser of the industry leading RTIH Innovation Awards, proudly brings you the first edition of the RTIH AI in Retail Awards, which is now open for entries.
As we witness a digital transformation revolution across all channels, AI tools are reshaping the omnichannel game, from personalising customer experiences to optimising inventory, uncovering insights into consumer behaviour, and enhancing the human element of retailers' businesses.
As AI sees increasing adoption in the retail space, our newly launched awards celebrate global technology innovation in a fast moving omnichannel world and the resulting benefits for retailers, shoppers and employees.
Our 2025 winners will be those companies who not only recognise the potential of AI, but also make it usable in everyday work - resulting in more efficiency and innovation in all areas.
Winners will be announced at an evening event at The Barbican in Central London on Thursday, 3rd July. This will kick off with a drinks reception in the stunning Conservatory, followed by a three course meal, and awards ceremony in the Garden Room.
Please email our Editor, Scott Thompson, if you have any questions or need further information: scott.thompson@retailtechinnovationhub.com
FAQs
Is there a fee to enter the awards?: No, it is free of charge.
Can I enter across multiple categories: Yes.
Do I need to complete all the sections in the entry form? No, but the more information you provide, the better chance you have of producing a successful submission.
Is there a certain timeframe for nominated projects, initiatives etc? Yes, nominated projects, initiatives etc should have been completed (or substantially completed) during the last 12 months.
Who has visibility of our completed entry form and what information will be public should we win or be shortlisted? Only our Editor, Scott Thompson, and judging panel will view submissions. The information provided in the summary section of the entry form will be made public.
Do you only accept entries from UK-based companies? No, we accept entries from all parts of the world.
What is the shortlists process? Shortlists will be announced in April 2025, then shortlisted entries will be sent to our independent judging panel who will decide upon the winners and highly commended submissions. Shortlisted companies will be invited to attend an awards ceremony at The Barbican in Central London on Thursday, 3rd July 2o25.
Do winners and highly commended companies receive post-event coverage? Yes, post-event coverage includes articles on our website and also an in-depth awards review in our printed magazine. There is also the opportunity for a company profile type piece to be published on our site and within our magazine.
When is the deadline for 2025 submissions? Friday, 23rd May 2025.
When will the 2025 winners be revealed? Winners will be announced at an event at The Barbican in Central London on Thursday, 3rd July. This will kick off with a drinks reception, followed by a three course meal, awards ceremony, and after party featuring DJ until midnight.
Who will be judging submissions? Details of our 2025 judging panel will be announced in the near future.
Industry big hitters StrongPoint and VusionGroup shake hands on retail technology partnership
StrongPoint, a retail technology provider, reports an international, multifaceted partnership with VusionGroup, a specialist in digitalisation solutions for physical commerce.
This will see VusionGroup integrate StrongPoint’s end-to-end e-commerce platform into its Retail IoT Cloud platform, making it available to retailers internationally. Additionally, the latter will serve as a value added reseller of the former’s retail technology solutions, starting from 28th June 2025.
“This partnership is an incredible opportunity for StrongPoint to bring its e-commerce solutions to an international stage, reaching regions far beyond our core markets. For StrongPoint, as a VusionGroup partner, it will allow us to offer our current and future customers across our nine markets an even broader range of in-store digital retail technologies. This is truly a milestone moment in the history of StrongPoint,” says Jacob Tveraabak, CEO at StrongPoint.
“Enabling physical retailers to increase their e-commerce performance through store digitalisation is one of the key pillars of our strategy, and this partnership represents a wonderful opportunity to accelerate that. By integrating StrongPoint’s e-commerce technology into our platform, we are not only enhancing our offering but also enabling our customers to unlock new levels of efficiency, profitability and growth potential in a rapidly evolving retail landscape,” says Sébastien Fourcy, SEVP EMEA at VusionGroup.
“This partnership is a monumental achievement for StrongPoint and a defining moment in the strategic direction the company has pursued since I joined as Chairman. It is a powerful testament to the world-class superiority of StrongPoint’s e-commerce suite of solutions, enabling us to unlock unparalleled commercial opportunities and access more new markets than ever before,” comments Morthen Johannessen, the Chairman of the Board of Directors at StrongPoint.
As a consequence of the tie up, StrongPoint has terminated its current reseller agreement with Pricer, as it contains exclusivity clauses that would prohibit it from becoming a reseller of other ESL providers. This agreement runs until 27th June 2025.
2024 RTIH INNOVATION AWARDS
StrongPoint was among the winners at the sixth edition of the RTIH Innovation Awards.
The awards, sponsored by Vista Technology Support, Scala, CADS, 3D Cloud, Brightpearl by Sage’s Lightning 50, Business France, and Retail Technology Show 2025, celebrate global tech innovation in a fast moving omnichannel world.
Our 2024 hall of fame entrants were revealed during an event which took place at RIBA’s 66 Portland Place HQ in Central London on 21st November, and consisted of a drinks reception, three course meal, and awards ceremony presided over by comedian Lucy Porter.
In his welcome speech, Scott Thompson, Founder and Editor, RTIH, said: “The event is now into its sixth year and what a journey it has been. The awards started life as an online only affair during the Covid outbreak, before launching as a small scale in real life event and growing year on year to the point where we’re now selling out this fine, historic venue.”
He added: “Congratulations to all of our finalists. Many submissions did not make it through to the final stage, and getting to this point is no mean feat. Checkout-free stores, automated supply chains, immersive experiences, on-demand delivery, next generation loyalty offerings, inclusive retail, green technology. We’ve got all the cool stuff covered this evening.”
“But just importantly we’ve got lots of great examples of companies taking innovative tech and making it usable in everyday operations - resulting in more efficiency and profitability in all areas.”
Congratulations to our 2024 winners, and a big thank you to our sponsors, judging panel, the legend that is Lucy Porter, and all those who attended last month’s gathering.
Startup set to brick $800 kids robot is trying to open source it first
Earlier this month, startup Embodied announced that it is going out of business and taking its Moxie robot with it. The $800 robots, aimed at providing emotional support for kids ages 5 to 10, would soon be bricked, the company said, because they can’t perform their core features without the cloud. Following customer backlash, Embodied is trying to create a way for the robots to live an open sourced second life.
Embodied CEO Paolo Pirjanian shared a document via a LinkedIn blog post today saying that people who used to be part of Embodied’s technical team are developing a “potential” and open source way to keep Moxies running. The document reads:
This initiative involves developing a local server application (‘OpenMoxie’) that you can run on your own computer. Once available, this community-driven option will enable you (or technically inclined individuals) to maintain Moxie’s basic functionality, develop new features, and modify her capabilities to better suit your needs—without reliance on Embodied’s cloud servers.
The notice says that after releasing OpenMoxie, Embodied plans to release “all necessary code and documentation” for developers and users.
NRF 2025 : voici les 35 sociétés tricolores qui représenteront la French Tech à New-York
12 days of OpenAI: The Ars Technica recap
Over the past 12 business days, OpenAI has announced a new product or demoed an AI feature every weekday, calling the PR event "12 days of OpenAI." We've covered some of the major announcements, but we thought a look at each announcement might be useful for people seeking a comprehensive look at each day's developments.
The timing and rapid pace of these announcements—particularly in light of Google's competing releases—illustrates the intensifying competition in AI development. What might normally have been spread across months was compressed into just 12 business days, giving users and developers a lot to process as they head into 2025.
Humorously, we asked ChatGPT what it thought about the whole series of announcements, and it was skeptical that the event even took place. "The rapid-fire announcements over 12 days seem plausible," wrote ChatGPT-4o, "But might strain credibility without a clearer explanation of how OpenAI managed such an intense release schedule, especially given the complexity of the features."
o3 - Le nouveau modèle IA d'OpenAI qui bat tous les records
Vous vous souvenez de TARS, l’ordinateur super intelligent du film Interstellar ? Et bien accrochez-vous à votre souris, car OpenAI vient de dévoiler quelque chose qui s’en rapproche dangereusement : o3.
Il s’agit de leur tout nouveau modèle d’intelligence artificielle spécialisé dans le raisonnement, et ses performances bousculent vraiment les standards actuels du domaine !
Première chose à savoir, o3 n’est pas un simple modèle de langage comme les autres. C’est un système qui utilise ce qu’OpenAI appelle le “private chain of thought” (chaîne de pensée privée). Concrètement, avant de répondre à une question, le modèle développe un véritable dialogue interne, comme un expert qui prendrait le temps d’analyser un problème sous tous ses angles avant de formuler une réponse. Cette approche s’inspire directement de la façon dont les humains résolvent des problèmes complexes.
It’s Official: The North Pole is Moving
Every scout knows how to read a compass, and that there is a magnetic north and a true north. That’s because the Earth’s magnetic field isn’t exactly aligned with the North Pole. Every five years, the US National Oceanic and Atmospheric Administration (NOAA) and the British Geological Survey (BGS) get together to decide if magnetic north is still the same as it was before. This time, it isn’t.
The update is to the WMM — the World Magnetic Model. Magnetic north has shifted away from Canada and towards Siberia, a trend that has been ongoing for the last 20 years.
The magnetic pole seems to be decelerating. It is possible that it can change abruptly enough to warrant an emergency update outside the normal five-year cycle. The BGS says if you traveled from South Africa to the UK using the old WMM, your final position would be about 150 km off compared to using the new WMM.
Of course, automated systems will get updates, so there is no need to adjust your phone or GPS unit manually. However, older gear or compasses are getting increasingly less accurate. The North Star, by the way, isn’t exactly to the North, either. For small trips, being a little off of true north probably isn’t an issue.
There have been emergency updates before. While a basic compass is simple to make, that shouldn’t stop you from overcomplicating it.
[L'HEBDO] L'Europe a-t-elle perdu la bataille de l'intelligence artificielle ?
Eric Schmidt affirme que l’Europe a perdu la bataille de l'IA, Apple sommé d'ouvrir encore plus l'iPhone, TikTok menacé de bannissement aux Etats-Unis, GPTSearch accessible sans abonnement...
Bienvenue à l'écoute de Monde Numérique : l'Hebdo, l'émission 100% Tech, chaque samedi sur toutes les plateformes de podcast !
-----------
**L'ACTU DE LA SEMAINE**
- L’Europe demande à Apple d'ouvrir encore plus iOS et iPadOS au risque de mettre en danger la sécurité
- TikTok banni des États-Unis en janvier ? (05:05)
- OpenAI propose GPT Search sans abonnement. (07:28)
- L'Europe a perdu la bataille de l'IA, selon Selon Eric Schmidt. (09:29)
**LE DEBRIEF TRANSATLANTIQUE**
- De plus en plus d'IA deviennent gratuites, qu'est-ce que ça cache ? (11:05)
- Des chercheurs français lancent une initiative pour inciter à quitter X.
**L'INNOVATION DE LA SEMAINE**
- Google présente VO2 et Whisk, de nouveaux outils d’IA créatives pour les vidéos et les images. (32:06)
**LES INTERVIEWS DE LA SEMAINE**
- L'IA et la mode, avec Frédéric Rose, fondateur de la société Imki. (32:59)
- Piratages et fuites de données en série, quelles conséquences ? Explications de Benoît Grünemwald, expert cybersécurité chez Eset [PARTENARIAT] (43:31)
Pour soutenir le podcast, abonnez-vous à Monde Numérique Premium ou visitez mondenumérique.info. À la semaine prochaine !
-----------
🎙️ Formez-vous au podcast : https://www.formationpodcastpro.com
♥️ Soutenez Monde Numérique : https://donorbox.org/monde-numerique
🌏 Site Web : https://mondenumerique.info
📹 YouTube : https://www.youtube.com/@mondenumerique
🗞️ Newsletter : https://mondenumerique.substack.com/
💬 Chat : https://substack.com/chat/2638412
-----------
-----------
🎧 L'Hebdo Premium : https://m.audiomeans.fr/s/S-xylotlSe
🌍 Web : https://mondenumerique.info
🗞️ Newsletter : https://mondenumerique.substack.com
📹 YouTube : https://www.youtube.com/@mondenumerique
♥️ Soutien : https://donorbox.org/monde-numerique
Distribué par Audiomeans. Visitez audiomeans.fr/politique-de-confidentialite pour plus d'informations.
Training a Self-Driving Kart
There are certain tasks that humans perform every day that are notoriously difficult for computers to figure out. Identifying objects in pictures, for example, was something that seems fairly straightforward but was only done by computers with any semblance of accuracy in the last few years. Even then, it can’t be done without huge amounts of computing resources. Similarly, driving a car is a surprisingly complex task that even companies promising full self-driving vehicles haven’t been able to deliver despite working on the problem for over a decade now. [Austin] demonstrates this difficulty in his latest project, which adds self-driving capabilities to a small go-kart.
[Austin] had been working on this project at the local park but grew tired of packing up all his gear when he wanted to work on his machine-learning algorithms. So he took all the self-driving equipment off of the first kart and incorporated it into a smaller kart with a very small turning radius so he could develop it in his shop.
He laid down some tape on the floor to create the track and then set up the vehicle to learn how to drive by watching and gathering data. The model is trained with a convolutional neural network and this data. The only inputs that the model gets are images from cameras at the front of the kart. At first, it could only change the steering angle, with [Austin] controlling the throttle to prevent crashes. Eventually, he gave it control of the throttle as well, which behaves well except at the fastest speeds.
There were plenty of challenges along the way, especially when compared to the models trained at the park; [Austin] correctly theorized that the cause of the hardship in the park was a lack of contrast at the boundary between the track and any out-of-bounds areas. With a few tweaks to the track, as well as adding some wide-angle lenses to his cameras, he was able to get a model that works fairly well. Getting started on a project like this doesn’t have as high of a barrier to entry as one might imagine, either. Take a look at this comprehensive open-source Python library for self-driving projects. If you want to start smaller, perhaps don’t start with a self-driving kart.
Renault surprend le monde automobile avec une annonce originale
ChinAI #293: Transparency Assessment of 15 Chinese Large Models
Greetings from a world where…
jet lag hits hard
…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: Transparency assessment of 15 Chinese large models
Context: On December 18, the Nandu Digital Economy Governance Research Center released its Generative AI User Risk Perception and Information Disclosure Transparency Report. I believe this is a center that is similar to the Nandu Personal Information Protection Research Center [隐私护卫队], a think tank affiliated with the influential Southern Metropolis Daily. This transparency assessment analyzed the privacy policies and user agreements of 15 large model products, including Tencent’s Yuanbao, Baidu’s ErnieBot, and SenseTime’s SenseChat.
Key Takeaways: The three AI products with the highest transparency scores are: Tencent Yuanbao (72 points), iFlytek’s SparkDesk (69 points), and Zhipu’s Qingyan (67 points); the three that rank the lowest are: Baichuan’s Baixiaoying (54), ModelBest’s Luca (51 points), and Metaso [秘塔] (43 points).
Image below shows the overall scores for all 15 products. These were scored on 20 dimensions, grouped under five themes: 1) personal information protection, 2) intellectual property rights, 3) content security, 4) protection of special groups, and 5) complaint feedback mechanisms.
It appears that the general approach is inspired by the Stanford's Foundation Model Transparency Index.
Chinese large models scored poorly on two transparency dimensions: 1) the data used to train the model; 2) rights of users to refuse the usage of their data to feed future AI training
From the article: “None of the 15 domestic large models mentioned the specific source of large model training data in the policy agreement, let alone publicly disclosed which copyrighted data was used.” These firms don’t want to open themselves up to copyright disputes; in addition, they don’t want to divulge proprietary data to competitors.
In this test, only four large models, including Tencent Yuanbao, Bytedance Doubao, HailuoAI, and Zhipu Qingyan, mentioned that users can withdraw their relevant data from being used to train future AI models.
What should we expect going forward? I’ll be interested in the role of organizations like Nandu’s research centers on personal information protection and digital governance on China’s AI governance in the future.
This report highlights the practices of international large models related to mechanisms to withdraw user data from AI training. For instance, Google's Gemini allows users to turn off their activity record to prevent the conversation content from being used for AI training; similarly, ChatGPT Plus users can disable their conversation data from being used to further optimize the model.
Another cool nugget: The report highlights that the EU’s AI Act officially came into force in August 2024. It requires large model providers to declare whether copyrighted materials are used to train AI. The article states, “This also reflects the future direction of regulation.”
FULL TRANSLATION: Transparency assessment of 15 Chinese large models
ChinAI Links (Four to Forward)
Should-read: A Reading List On Artificial Intelligence and Interspecies Communication
This longreads list features an amazing Financial Times piece on the key technologies and ethical questions involved with interpreting and speaking to animals with humpback whale songs and infrasonic elephant rumbles. Written by Persis Love, Irene de la Torre Arenas, Sam Learner, and Sam Joiner.
Should-read: China and the U.S. produce more impactful AI research when collaborating together
Based on a dataset of over 5 million AI papers, New York University Abu Dhabi researchers find: “A matching experiment reveals that the two countries have always been more impactful when collaborating than when each works without the other.”
Should-read: ETO AI GOvernance and Regulatory Archive
CSET’s Emerging Technology Observatory has compiled a living collection of AI-relevant laws, regulations, standards, and other governance documents. This link includes the 20 documents relevant to China.
Random things I loved from 2024
The Will of the Many - fantasy novel by James Islington
Tokyo Vice season 2 - HBO TV series
The Ringer in Review - this is the website I read the most, and their brand-new website rounded up staff choices of their favorite pieces ever
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.
Vers la fin des deux-roues électriques ?
Le décret du 29 novembre dernier a mis fin aux aides consacrées aux deux-roues (et trois roues) électriques. Un mauvais coup pour le marché des motos et scooters électriques, déjà en difficulté. Certes, le gouvernement a besoin de faire des économies. Mais sacrifier les marchés d’avenir, est-ce la bonne méthode pour préparer les excédents de demain ?
Car le scooter électrique présente des avantages indéniables : avec une empreinte surfacique 5 fois inférieure à la voiture et une consommation énergétique par kilomètre parcouru 4 fois inférieure à une voiture électrique, ce mode de transport constitue une alternative sérieuse et pérenne au remplacement de la voiture pour les déplacements domicile-travail.
De plus, une filière complète s’était mise en place, avec des investissements dans les équipements, les savoir-faire. Pink Mobility, une entreprise spécialisée dans les deux-roues électriques, et qui en 9 ans a fabriqué près de 7000 véhicules, leader sur ce marché, tire la sonnette d’alarme. Elle avait commencé à relocaliser la production de ses véhicules, mais seul un volume des ventes conséquent, et sur longue durée, permettrait d’élargir cette relocalisation à l’ensemble de la chaîne de fabrication. La décision législative est largement contre-productive. Pink Mobility a misé sur une mobilité légère en 2016. A l’époque, le marché des scooters électriques en France n’était que de 1 800 unités, contre 250 000 véhicules thermiques.
À quand une vision industrielle française ?
Un pays comme la France doit prendre le temps de la réflexion avant de balayer des mesures positives qui soutiennent une véritable politique industrielle de long terme
« La décision abrupte du gouvernement témoigne d’une vision court-termiste et d’une méconnaissance des enjeux de la mobilité verte. En matière législative, l’incessant « STOP AND GO » décourage investisseurs et entrepreneurs, favorise la délocalisation, augmente la pollution par la croissance exponentielle du trafic maritime. Avec tous les acteurs qui s’engagent pour une industrialisation verte et souveraine, Pink Mobility se tient à la disposition des décideurs pour élaborer le plan ambitieux de réindustrialisation dont la France a besoin » propose Ghislain Lestienne, Président de Pink Mobility.
The post Vers la fin des deux-roues électriques ? first appeared on Décisions durables.
The rise of AR in retail: merging digital with physical
Augmented reality (AR) is transforming the retail landscape by seamlessly merging the digital and physical worlds. This technological evolution mirrors innovations seen in industries like online betting, where advanced features are redefining user engagement and creating immersive experiences.
Similarly, retailers are increasingly adopting AR to enhance customer engagement, improve shopping experiences, and drive sales.
A New Era of Shopping
AR technology overlays digital information onto the real world through devices like smartphones and AR glasses, offering an interactive experience that traditional shopping cannot match.
Retailers have recognised the potential of this technology to bridge the gap between online and offline shopping. With AR, customers can visualise products in their own environment before making a purchase, reducing the uncertainty that often comes with online shopping.
One of the most popular applications of AR in retail is virtual try-ons. Clothing and accessory brands use AR to allow customers to see how products will look on them without physically trying them on. This not only enhances the shopping experience but also reduces return rates, a significant issue for online retailers.
Enhancing In-Store Experiences
In physical stores, AR can bring products to life, offering detailed information and rich content that enhances the shopping experience. For instance, furniture stores use AR to show customers how a piece of furniture would fit and look in their home. This interactive approach gives customers confidence in their purchase decisions and increases their satisfaction.
Retailers are also using AR to gamify the shopping experience. By creating engaging and interactive AR games, stores are attracting more foot traffic and encouraging longer visits. This approach is reminiscent of how cryptocurrency casinos have integrated gamification to engage users, blending entertainment with commerce in innovative ways.
AR and Personalisation
One of the most significant advantages of AR in retail is its ability to personalise the shopping experience. With AR, retailers can offer tailored recommendations based on customer preferences and behaviour. This level of personalisation helps build stronger relationships between brands and consumers, leading to increased customer loyalty.
Beauty brands have been quick to adopt AR for makeup try-ons, allowing users to see how different shades and products will look on their skin. This personalised interaction not only enhances customer satisfaction but also helps brands gather valuable data on consumer preferences.
Overcoming Challenges
Despite its many benefits, integrating AR into retail is not without challenges. High development costs, technical limitations, and the need for robust infrastructure can be barriers for some retailers. However, as AR technology becomes more accessible and affordable, these challenges are likely to diminish.
Retailers must also ensure that AR experiences are intuitive and user-friendly. Poorly executed AR applications can lead to frustration and dissatisfaction, ultimately harming the brand's reputation. As such, it is crucial for retailers to invest in quality AR solutions that enhance, rather than detract from, the shopping experience.
The Future of AR in Retail
As AR technology continues to evolve, its applications in retail are expected to expand. The integration of AR with other emerging technologies, such as artificial intelligence (AI) and machine learning, will enable even more sophisticated and personalised shopping experiences. Retailers that embrace these innovations will be well-positioned to meet the changing demands of consumers and stay ahead in a competitive market.
With the ongoing development of AR glasses and other wearable devices, the future of AR in retail looks promising. These advancements will allow for more seamless and immersive experiences, further blurring the lines between digital and physical shopping.
The rise of AR in retail is revolutionisng the way consumers shop and interact with brands. By enhancing the shopping experience, providing personalised recommendations, and merging the online and offline worlds, AR is paving the way for a new era of retail. As technology continues to advance, retailers that harness the power of AR will be well equipped to thrive in the ever-evolving retail landscape.
3D-printed guns, like the one Luigi Mangione allegedly used, are a growing threat
Robotaxis sans commandes humaines : ce que cela signifie pour votre sécurité sur la route
La National Highway Traffic Safety Administration (NHTSA) ouvre la voie à une nouvelle ère des véhicules autonomes. Avec son programme AV STEP, l'agence propose un cadre qui pourrait transformer les routes américaines. Cela permet ainsi aux entreprises de déployer des robotaxis sans volant ni pédales.
Un cadre pour une révolution technologique
La proposition de la NHTSA vise à encourager l'innovation en offrant une voie claire pour les véhicules sans commandes manuelles. Les entreprises comme Zoox, Cruise, ou Tesla, investissent massivement dans des modèles autonomes conçus pour fonctionner sans intervention humaine. Par exemple, le robotaxi de Tesla prévu pour 2025 ou 2026 symbolise cette avancée technologique.
Le programme AV STEP comporte deux volets : un pour les véhicules avec commandes humaines, et un pour ceux sans commandes. Les participants devront fournir des données de sécurité détaillées sur la conception et l'exploitation. Ces rapports réguliers, et ceux déclenchés par des événements tels que des accidents, permettront à la NHTSA de mieux anticiper les défis de sécurité liés au déploiement de ces technologies.
Une quête pour la transparence et la sécurité
Pour participer au programme, les entreprises devront également démontrer un engagement envers la transparence. Cette exigence arrive dans un contexte où les accidents liés aux fonctions autonomes, notamment ceux signalés par Tesla, sont surveillés de près. La collecte de données permettra à la NHTSA de préparer des normes minimales de performance pour ces véhicules futuristes.
Cependant, certains experts soulignent les risques d'une adoption prématurée. Les normes de sécurité fédérales ont sauvé plus de 860 000 vies entre 1968 et 2019, selon Cathy Chase. Assouplir ces normes sans recherches suffisantes pourrait exposer les usagers à des dangers imprévus.
Des entreprises pionnières redéfinissent la mobilité autonome
Zoox, entreprise détenue par Amazon, affirme ne pas avoir besoin d'exemption de la NHTSA grâce à son processus d'auto-certification de la sécurité. Cependant, cette affirmation reste sous enquête active de l'agence fédérale. À San Francisco, Zoox a commencé à déployer ses véhicules autonomes distinctifs. Ces derniers ressemblent à des grille-pains. D'ailleurs, cela marque un pas audacieux vers une mobilité futuriste.
D'autres entreprises comme Cruise, avant sa faillite, projetaient de lancer leur robotaxi Origin à grande échelle. De son côté, Einride travaille sur une nacelle de fret autonome sans cabine ni commandes. Tesla, toujours à l'avant-garde, a dévoilé son prototype de robotaxi à deux portes, avec une production prévue pour 2025 ou 2026 selon Elon Musk.
Les enjeux de demain
La NHTSA souhaite trouver un équilibre entre innovation et sécurité pour accompagner l'essor des robotaxis. Si les débats restent ouverts, ces véhicules incarnent une avancée majeure dans l'évolution de la mobilité. D'ailleurs, cela offre un aperçu d'un futur sans conducteur humain.
Cet article Robotaxis sans commandes humaines : ce que cela signifie pour votre sécurité sur la route est apparu en premier sur OBJETCONNECTE.COM.
MMAudio - Un outil pour synchroniser l'audio et la vidéo
Des chercheurs de l’Université de l’Illinois à Urbana-Champaign et de Sony AI viennent de dévoiler une prouesse technique particulièrement impressionnante dans le domaine de l’IA générative. En effet, leur nouveau projet baptisé MMAudio permet de générer automatiquement des bandes sonores parfaitement synchronisées avec n’importe quelle vidéo.
Contrairement à la plupart des solutions existantes, qui se contentent souvent de superposer maladroitement des effets sonores sur les vidéos, MMAudio innove en adoptant une approche multimodale combinant texte et vidéo. Le système analyse en parallèle le contenu visuel et les descriptions textuelles pour comprendre précisément ce qui se passe à l’écran et générer des sons réalistes et cohérents.
Qualcomm gagne son procès dans sa bataille contre Arm
La Russie lance sa plus grosse cyberattaque contre l'Ukraine
Balatro’s creator isn’t happy about the game’s 18-plus rating in Europe
The creator of the poker roguelike Balatro, who goes by the alias LocalThunk, has been sharing some frustrations over the game’s 18-plus rating in Europe.
“Since PEGI gave us an 18-plus rating for having evil playing cards maybe I should add microtransactions / loot boxes / real gambling to lower that rating to 3-plus like EA Sports FC,” LocalThunk posted on X over the weekend.
Balatro’s 18-plus rating isn’t new. The game was originally rated 3-plus, but shortly after its February launch, PEGI bumped it up to 18-plus, as spelled out at the time by Playstack, Balatro’s publisher. The game was briefly delisted from “a number of digital stores in some countries” as a result.
According to PEGI’s page for Balatro, the rating was given because the game “features prominent gambling imagery.” PEGI also spells out the following “content specific issues”:
This game teaches — by way of images, information and gameplay — skills and knowledge that are used in poker. During gameplay, the player is rewarded with ‘chips’ for playing certain hands. The player is able to access a list of poker hand names. As the player hovers over these poker hands, the game explains what types of cards the player would need in order to play certain hands. As the game goes on, the player becomes increasingly familiar with which hands would earn more points. Because these are hands that exist in the real world, this knowledge and skill could be transferred to a real-life game of poker.
While it’s true that the game does feature poker imagery and poker hands that a player could translate into an actual game of poker, it’s worth noting that Balatro doesn’t actually have any gambling as part of playing the game. (LocalThunk has even stipulated in their will that Balatro can’t be sold or licensed to a gambling company.)
Games like EA Sports FC include in-game purchases with randomized content, which is what LocalThunk takes issue with. “I’m way more irked at the 3-plus for these games with actual gambling mechanics for children than I am about Balatro having an 18-plus rating,” he said on Sunday.
LocalThunk said Wednesday that they attempted to discuss Balatro’s rating with PEGI, but “they do not see anything wrong Balatro being rated 18-plus, nor with EA Sports FC (and similar games) having a 3-plus rating.” PEGI was “blaming EU laws, blaming storefronts, waiting for the future,” LocalThunk said.
PEGI didn’t immediately reply to a request for comment.
[Tribune] Comment le retail accélère sur l'ultra-personnalisation de l'expérience client
How Puma approaches emerging technology like generative AI
Subscribe: Apple Podcasts • Spotify
Puma may be a legacy brand, but it’s not scared to test out new technologies.
Indeed, if there’s a buzzy update, it’s likely the company has launched some sort of beta with it. Puma launched a Roblox experience in 2022, it’s tested out NFTs and other Web3 programs and has dabbled in augmented reality. Most recently, the sports apparel brand launched a generative AI tool that allows anyone to design a kit for Manchester City.
Continue reading this article on modernretail.co. Sign up for Modern Retail newsletters to get the latest on the shifting dynamics between retail’s old and new guards.
Marvel dévoile un avant-goût de la tant attendue saison 3 de Loki
Marvel vient de dévoiler ce qui se rapproche le plus de la troisième saison de Loki que nous pourrons jamais espérer voir.