🚀 Best robot videos, pigeon cyborg, fusion homme-machine & autres news du futur

Chien robot se rebelle, vie sur mars ? et plus

Bonjour,

Vous recevez la version gratuite de la newsletter Parlons Futur : une fois par semaine au plus, une sélection de news, mêlant sources anglophones et francophones, résumées en bullet points sur des sujets tech 🤖, science 🔬, éco 💰, géopolitique 🌏, défense ⚔️ et espace 🚀 pour mieux appréhender le futur 🔮.

Je m’appelle Thomas, plus d’infos sur moi en cliquant ici.

🎙️ Mes derniers podcasts

• Poutine et la guerre à l’Europe : le scénario catastrophe du front atlantique

  • Entretien exceptionnel (Spotify, Apple, Youtube) avec Stéphane Audrand, consultant indépendant en risques internationaux, historien et officier de réserve dans la Marine. Il est l’auteur remarqué de l’article “Poutine et la guerre à l’Europe : le scénario du front atlantique” publié dans le Grand Continent.

• Vivons-nous dans une simulation ? (Youtube)

• 🌍Entretien exclusif avec Philippe Bihouix, co-auteur de la BD à succès “Ressources : Un défi pour l’humanité” (Youtube, Spotify, Apple)

Et voici donc ma toute dernière sélection !

• Yann LeCun Raises $1 Billion to Build AI That Understands the Physical World (Wired)

  • a new Paris-based startup!

  • LeCun argues that most human reasoning is grounded in the physical world, not language, and that AI world models are necessary to develop true human-level intelligence. “The idea that you’re going to extend the capabilities of LLMs to the point that they’re going to have human-level intelligence is complete nonsense”.

  • AMI aims to build “a new breed of AI systems that understand the world, have persistent memory, can reason and plan, and are controllable and safe”

• Five Million Robots, Six Months, One Manhattan

  • The Mid-Journey founder dropped a thought experiment:

  • “5 million humanoids working 24/7 can build Manhattan in six months. Imagine what the world looks like when you have 10 billion of them by 2045.”

  • If that’s even remotely true—and the math seems to check out—the implications are staggering

• Impressive : a Chinese robot is made to learn playing tennis quite well in a short time (voir la vidéo)

  • Comparaison humaine : un débutant adulte a typiquement besoin de 20–40 heures de cours et pratique pour atteindre un taux de retour régulier de ~80% en coup droit.

  • Le robot y parvient avec 5 heures de données de motion capture, provenant de 5 joueurs amateurs exécutant des mouvements de base dans un espace de seulement 3m × 5m, plus quelques heures de calcul GPU.

  • Mais attention au périmètre : le robot ne sert pas, ne joue pas de volées, ne gère qu’un demi-court, et la raquette est fixée au bras par un adaptateur 3D, contact balle-raquette de quelques millisecondes sans contrôle fin des effets.

• 😂 A robot in California just smashed some dishes started dancing instead of working(voir la vidéo)

  • Staffers at the restaurant were forced to intervene. One employee can be seen holding the robot by the scruff of its neck while navigating her phone, presumably searching the bot’s app for controls.

  • In the end, it took three workers to restrain the robot, who were all forced to duck periodically to avoid being smacked by its waving and sauce-covered hands.

• 😂 Robot in China Escorted Away By Cops After Terrorizing Old Woman (voir la vidéo)

• An LLM-controlled robot dog saw us press its shutdown button, and the LLM rewrote the robot’s code so it could stay on. (Palisade Research)

• US General PETRAEUS: The future of warfare—not yet seen in the Gulf—is Ukraine producing 7 million drones per year (source)

  • Last year they made 3.5 million, enabling 9–10 thousand drones per day.

  • These are still remotely piloted. But autonomous systems are coming, and that means drone swarms. They’re already being introduced at mission end. Defending against swarms is really hard—you’ll need high-powered microwaves or similar.

  • We’re not where we should be, despite what we should have learned from Ukraine for a very long time. They’re making software changes every week or two, hardware changes every 2–3 weeks.

• What?? A Russian startup has turned pigeons into living drones using neural implants. (source)

  

  • the interface delivers “mild stimulation” to parts of the pigeon’s brain, prompting it to prefer certain flight paths—similar to how standard drones follow pre-loaded flight tasks—while the bird “otherwise behaves naturally.”

  • To turn the pigeons into living drones, company technicians insert tiny electrodes through the birds’ skulls, then connect the electrodes to a stimulator mounted on the pigeons’ heads.

  • However, Neiry has yet to provide independent reviews from third-party sources, leading some experts to question the ethical implications of their technology.

• Human Brain Cells on a Chip Learned to Play Doom in a Week (New Scientist)

  • “A clump of human brain cells can play the classic computer game Doom. While its performance is not up to par with humans, experts say it brings biological computers a step closer to useful real-world applications, like controlling robot arms.”

• Recent 3D Printing Breakthrough (source)

  • “A team at [MIT] has recently developed a printer with four different extruders that outputs five different materials to produce a fully functioning linear motor in about three hours. ...The team explained how by retrofitting a printer with enough extruders to handle the various materials needed to make a working motor, they decimated the usual production time for such a device and brought the material costs down to around $0.50.”

• C’est confirmé. La mission DART de la NASA a réussi pour la première fois en 2022 à modifier la trajectoire d’un astéroïde, démontrant qu’une défense planétaire est possible. (TTSO)

• Is AGENCY, and not simply INTELLIGENCE, what makes superintelligent AI so dangerous ?

  • Not necessarily, says Yann LeCun:

  • Danger does not come merely from agency, but from agency with no ability to anticipate consequences and with no safety guardrails.
    The solution?
    AI agents that can predict the consequences of their actions (world models) and only take actions whose predicted outcomes satisfy safety guardrails.
    I’ve been saying this for over 5 years.
    I’ve been designing objective-driven AI systems based on world models for that reason.

• Pour certaines figures de la tech comme Peter Thiel, cofondateur de PayPal, il n’est plus nécessaire de faire des études longues…

  • Yann LeCun répond dans Challenges:

  • “C’est absolument faux. Le progrès technologique se base de plus en plus sur la science. Il faut au contraire des ingénieurs encore plus qualifiés pour piloter l’IA et innover. Il n’y a pas de chemin détourné mais il faut repenser les cours afin que les étudiants apprennent à apprendre. Car les techniques évoluent très vite.”

• Les experts en santé mentale d’OpenAI s’opposent unanimement au lancement du mode érotique de ChatGPT (source)

• Anthropic recrute un expert en armes chimiques et explosifs pour renforcer les garde-fous de son IA Claude (BBC)

• Anthropic is taking legal action against the US Department of Defence (BBC)

  • After the US Department of Defence designated it a supply chain risk when the firm insisted its systems must not be used in either fully autonomous weapons or mass surveillance of Americans.

  • Anthropic co-founder Dario Amodei wrote in February that he didn’t think the technology was good enough yet, and should not be used for these purposes.

  • The White House said the US military would not be governed by tech companies.

  • The risk label puts the US company in the same boat as the Chinese telecoms firm Huawei, which was similarly blacklisted over different national security concerns.

• La société d’Elon Musk xAI (qui fait maitenant partie de SPaceX) perd 9 de ses 11 cofondateurs en 3 ans et entame une restructuration profonde (Techcrunch)

  • “xAI was not built right first time around, so is being rebuilt from the foundations up,” Musk said.

• Sam Altman on Putting Data Centers in Space: ‘Ridiculous for Now’ (source)

• Brain Implants Let Paralyzed People Type Nearly as Fast as Smartphone Users (source)

• An AI disaster is getting ever closer (The Economist)

  • Another realm in which ai is making alarming leaps is the development of biological and chemical weapons, which, OpenAI warned in August, creates a “significantly increased likelihood and frequency of biological or chemical terror events”.

  • Firms that manufacture DNA to order have long been able to check customers’ requests against databases of dangerous genes, making it hard to, say, create a genetically engineered bacterium to produce ricin, a neurotoxin.

  • But in October in a study published in Science, researchers from Microsoft and IBBIS, an international biosafety group, pointed out that improvements in AI-powered protein design make it possible to create genes to produce an analogous toxin that share no DNA with the original gene.

  • Requests to buy such AI-redesigned sequences would not be detected by existing vetting systems, although the researchers did propose a fix.

• New AI Video Generator Is So Impressive That It’s Scaring Hollywood “I hate to say it. It’s likely over for us. (source)

  • Crazy example 1, example 2, example 3.

• Google: the median prompt for Google’s Gemini AI model uses the same amount of energy it takes to power 9 seconds of television and consumes around five drops of water (source)

• Did you know that chimpanzees actually have a better short term memory than humans? (source)

  • Sur ce thème, sur mon podcast : 🧠Contre-intuitif : pourquoi l’intelligence humaine baisse en fait depuis 10,000 ans (Youtube, Spotify, Apple)

• NASA Running Out of Non-Life Explanations for What Its Rover Found on Mars: long-chain organic molecules (NASA)

  • A new study suggests that non-biological sources cannot fully account for the abundance of organic compounds found in a sample collected by the rover.

• A massive solar farm in the Gobi Desert created a new ecosystem (video)

  • Vegetation jumped from virtually 0% to 80% within 3 years!

  • Sheep were introduced to control the growth. That flock has grown to 20,000!

• JDN Tutoriel OpenClaw : installez et configurez votre premier agent IA autonome (source)

• Comment nous avons créé une vidéo publicitaire avec Kling (JDN)

Partagez cette newsletter par Whatsapp en cliquant ici ❤️

• Si vous appréciez cette première moitité de l’apéro gratuite, n’hésitez pas à prendre 3 secondes svp pour transférer cet email ou l’envoyer à ne serait-ce qu’un contact par Whatsapp en cliquant ici 🙂

• Et si on vous a partagé cet email, vous pouvez cliquer ici pour vous inscrire et ne pas manquer les prochains

👋👀🦉💡 Je vous invite maintenant à vous abonner à la version payante (l’équivalent du prix d’un café par semaine, le minium imposé par Substack) pour recevoir la semaine prochaine les résumés de ces longs articles issus des plus grands journaux (et accéder aux futures éditions payantes et archives) en cliquant sur le bouton “Upgrade to paid/Subscribe now” 👇

• 🔒 Palantir dévoile enfin comment son IA opère sur les théâtres de guerre

• 🔒 Why you shouldn’t dismiss Japanese robots!

• 🔒 When AI can undermine, rather than defend, democratic values : Two use cases should not be included in our contracts with the Department of War (Anthropic CEO Dario Amodei)

• 🔒 What if AI made the world’s economic growth explode?

Les précédentes newsletters

❤️ Transférez cette newsletter à vos amis, ou partagez-la par Whatsapp en cliquant ici.

Pour vous inscrire et ne pas manquer les prochaines éditions, vous pouvez cliquer ici.

Merci, et à bientôt !

Thomas