🚀 Révolution française en GoPro, décoder les émotions des cochons, battle d'IAs, chat mutant non-allergène & plus !
🚀 Révolution française en GoPro, décoder les émotions des cochons, battle d'IAs, chat mutant non-allergène & plus !
Photoshoper sa voix, robot gluant, the smartest bird ever, utérus artificiel surveillé par IA...
Bonjour,
Vous recevez la newsletter Parlons Futur : une fois par semaine au plus, une sélection de news résumées en bullet points sur des sujets tech 🤖, science 🔬, éco 💰, géopolitique 🌏 et défense ⚔️ pour mieux appréhender le futur 🔮.
Si on vous a envoyé cet email, vous pouvez vous inscrire ici pour ne pas manquer les prochains.
Cette newsletter est gratuite, si vous l’appréciez, prenez s’il vous plaît quelques secondes pour la transférer à 3 ami(e)s aujourd’hui, je vous en remercie par avance ! 🙂
Je m'appelle Thomas, plus d'infos sur moi en bas d'email.
Voici donc ma dernière sélection !
L’apéro
DeepMind is deciphering damaged ancient Greek texts (source)
Concept intéressant de réalité augmentée permettant de "compter les cartes" et évaluer les chances de gagner au poker (voir la vidéo de 25 secondes)
NVIDIA’s Tiny New AI Transforms Photos Into Full 3D Scenes in Mere Seconds (watch the 40-second video)
A new study from Stanford Medical School found that around the age of 13, children no longer find their mothers’ voices “uniquely rewarding.” (source)
Chinese technology: watch a swarm of drones autonomously track a human through a dense forest (1-min video)
This is going to result in some serious changes to military operations in wooded areas and jungle
Drôle, quand la police arrête une "self-driving car" : "Seeing the future collide with the present is so fun." (vidéo)
Hum... Musk says humanoid robots (first generation aimed for 2023) will be worth more than Tesla’s car business (source)
Amazing demo of a Japanese robot exoskeleton expanding a full human body (vidéo)
Despite their extremely futuristic appearance, skeletonics extend the movement of a human's arms and legs without the aid of any motor and it's not powered by electricity (source)
Watch the world's first electric wingsuit flight
US biofirm plans to make hypoallergenic cats using CRISPR gene editing (New Scientist)
The Eurasian magpie (la pie) is believed not only to be among the most intelligent of all animals. This one drinks from a bottle using water displacement by filling it with pebbles (cailloux) (30-second video)
Le Pentagone dispose d’un système laser pouvant identifier une personne à plus de 200 mètres de distance à partir d’une signature cardiaque (source)
À table !
DALL-E 2, le générateur d'images d'Open AI, est incroyable, aux humains d'être originaux avec leurs "prompts"
DALL-E — a portmanteau of the surrealist Salvador Dalí and Pixar’s WALL-E — takes text prompts and generates images from them. In January 2021, the company introduced the first version of the tool, which was limited to 256-by-256 pixel squares. (source)
But the second version, which entered a private research beta in April, feels like a radical leap forward. The images are now 1,024 by 1,024 pixels and can incorporate new techniques such as “inpainting” — replacing one or more elements of an image with another. (Imagine taking a photo of an orange in a bowl and replacing it with an apple.)
DALL-E has also improved at understanding the relationship between objects, which helps it depict increasingly fantastic scenes — a koala dunking a basketball, an astronaut riding a horse.
Initially DALL-E 2 would take around 15 seconds to generate 10 images, then the number of images was reduced to six, to allow more people access.
Some examples below of prompts shared with DALL-E and then the image it created in response:
"a medieval painting of the wifi not working" (source):
“A bear economist in front of a stock chart crashing, digital art”:
It’s striking the degree to which DALL-E captures emotion here: the fright and exasperation of the bear
DALL-E ne génère que des photos, mais qq a réussi à bricoler avec pour produire une vidéo où des tenues de mode sont générées par DALL-E et adpatées à une personne en train de marcher (voir la vidéo de 7 secondes), ça donne une idée des futurs outils et apps bientôt disponibles pour la mode !
Nouvelle feature annoncée le 1er septembre par le Psdt et Cofounder d'OpenAI: “Outpainting — extend an existing image, to arbitrary size:”
La battle d'AIs: qq a challengé les 3 principaux générateurs d'images du moment avec les mêmes requêtes :
"Beatles lego set, catalogue photograph"
Stablediffusion, un "concurrent" de DALL-E 2, n'est pas mal non plus, et offre un accès plus libre. Combiné à un autre outil, on peut soumettre un "prompt" sous la forme d'un dessin tout simple et le sublimer de façon bluffante grâce à l'IA, cela donne:
Midjourney, yet another tool, n’est pas en reste :
“GoPro footage of the French Revolution”:
Voir les autres photos incroyables ici (GoPro footage of Viking Raid, Battle of Trafalgar, D-Day landing, Fall of Berlin Wall)
Pour un bon récap, voir l'article La courte et folle histoire de l’art généré par intelligence artificielle (de DeepDream en 2015 à DALL·E 2 en 2022)
India is likely to be the world’s fastest-growing big economy this year (The Economist)
as the pandemic recedes, 3 pillars are clearly visible that will support growth in the next decade:
the forging of a single national market;
an expansion of industry owing to the renewable-energy shift and a move in supply chains away from China;
and a high-tech welfare safety-net for the hundreds of millions left behind by all this.
MIT’s Robotic Cheetah (guépard) Taught Itself How to Run and Set a New Speed Record in the Process (2-min video)
“In just three hours’ time, the robot experienced 100 days worth of virtual adventures over a diverse variety of terrains and learned countless new techniques for modifying its gait (démarche) so that it can still effectively loco-mote from point A to point B no matter what might be underfoot.
It might not be the most graceful thing to watch when moving at high speeds, but the Mini Cheetah hit a new top speed of 3.9 m/s, or a little over 14km/h, which is faster than the average human can run.”
Dingue: The Magnetic Slime Robot That Could Retrieve Accidentally Swallowed Objects (1-min video , source)
Researchers have developed a magnetic slime (substance gluante) "robot" that can shift into different shapes to grab objects. It can stretch out in multiple directions to grab items on opposite sides. It's self-healing, too.
It behaves like a liquid or solid depending on force, and can be controlled using external magnets. There are no robotics inside the slime at present, but you can steer it like a robot
We can now decode pigs’ emotions (Source)
Using thousands of acoustic recordings gathered throughout the lives of pigs, from their births to deaths, an international team of researchers is the first in the world to translate pig grunts into actual emotions across an extended number of conditions and life stages.
Using more than 7000 audio recordings of pigs, the researchers designed an algorithm that can decode whether an individual pig is experiencing a positive emotion (‘happy’ or ‘excited’), a negative one (‘scared’ or ‘stressed’) or somewhere in between
Des chercheurs chinois planchent sur la croissance des fœtus dans un utérus artificiel "surveillée par une intelligence artificielle" (source)
L'utérus artificiel, ou « appareil de culture d'embryon à long terme », est décrit comme un système contenant des fluides nutritifs dans lequel sont placés des embryons de souris. L'appareil est surveillé par une intelligence artificielle qui peut ajuster les flux de dioxyde de carbone et d'éléments nutritifs, et intervenir sur certains facteurs environnementaux.
Les chercheurs se limitent actuellement à des embryons d'animaux car la loi internationale limite la recherche sur les embryons humains à 14 jours de développement. Ils voudraient pourtant poursuivre au-delà car « il reste encore de nombreux mystères non résolus sur la physiologie du développement embryonnaire humain typique » selon Sun Haixuan, qui a dirigé la recherche.
Technology won’t be a problem for its future application, but legal and ethical concerns might, warns Beijing-based researcher
MIT develops a speaker (haut-parleur) thinner than sheet music (Techcrunch, 40-sec video)
ultra-thin loudspeakers large enough to cover the inside of an automobile or to wallpaper a room, whispering a promise of immersive sound without an obvious source.
This thin-film loudspeaker produces sound with minimal distortion while using a fraction of the energy required by a traditional loudspeaker. The hand-sized loudspeaker the team demonstrated, which weighs about as much as a dime, can generate high-quality sound no matter what surface the film is bonded to.
“It feels remarkable to take what looks like a slender sheet of paper, attach two clips to it, plug it into the headphone port of your computer and start hearing sounds emanating from it. It can be used anywhere.
“We have the ability to precisely generate mechanical motion of air by activating a physical surface that is scalable. The options of how to use this technology are limitless”
Bientôt le Photoshop pour la voix? (The Verge)
Sonantic, an AI voice startup, says it’s made a minor breakthrough in its development of audio deepfakes, creating a synthetic voice that can express subtleties like teasing and flirtation.
In this 1-minute video, you can hear the company’s attempt at a flirtatious AI. Company says the speech synthesized for its flirty video required “very little manual adjustment”
On a first listen, it seems the voice is near-indistinguishable from that of a real person
Sonantic CEO Zeena Qureshi describes the company’s software as “Photoshop for voice.” Its interface lets users type out the speech they want to synthesize, specify the mood of the delivery, and then select from a cast of AI voices, most of which are copied from real human actors.
Emotional choices for delivery include anger, fear, sadness, happiness, and joy, and, with this week’s update, flirtatious, coy, teasing, and boasting. A “director mode” allows for even more tweaking: the pitch of a voice can be adjusted, the intensity of delivery dialed up or down, and those little non-speech vocalizations like laughs and breaths inserted.
“Our clients are mostly triple-A game studios, entertainment studios, and we’re branching out into other industries. We recently did a partnership with Mercedes [to customize its in-car digital assistant] earlier this year.”
Forget Elon Musk's Neuralink: Paradromics is now the leader of high-bandwidth Brain Computer Interface, says Peter Diamandis, founder and chairman of the X Prize Foundation
"Up until today, I had been carefully watching Neuralink as the leading company in implantable BCI technology, but I’m now clear that Paradromics has taken the lead in the race to high-bandwidth BCI."
"by my evaluation, Paradromics is years ahead of Neuralink and has created an implantable system that can translate between bioelectric and digital signals, and act as a means to connect living brains with computational clouds."
"In the not-too-distant future, companies like Paradromics could potentially use BCIs to cure neuro-degenerative diseases, enhance memory, enhance meditation, improve sleep, or treat depression. After all, it’s all about your neurochemistry and once you can dial that up or down, it gives you command override."
Les dernières newsletters :
L’addition ?
Cette newsletter est gratuite, et si vous souhaitez m'encourager à continuer ce modeste travail de curation et de synthèse, je vous invite simplement à transférer cet email à 3 ami(e)s, merci d'avance :)
Quelques mots sur le cuistot
J'ai écrit plus de 50 articles ces dernières années, à retrouver ici, dont une bonne partie publiés dans des médias comme le Journal du Net (mes chroniques ici), le Huffington Post, L'Express, Les Échos.
Je suis CEO et co-fondateur de l'agence digitale KRDS, nous avons des bureaux dans 6 pays entre la France et l'Asie. Je suis basé à Singapour (mon Linkedin), également membre du think tank NXU.
Retrouvez-moi sur twitter en cliquant ici : je tweete des faits et infos contre-intuitives, brèves, à consommer sur place (et non pas des liens sans contexte vers des articles interminables), Jacques Attali est un de mes followers par exemple.
Retrouvez ici mon podcast Parlons Futur (ou taper "Parlons Futur" dans votre appli de podcast favorite), vous y trouverez entre autres des interviews et des résumés de livres.
Merci, et bon week-end !
Thomas