🚀 Ordi piloté avec langue; perroquet fait appel vidéo; machine lit nos pensées sans implant; selfie de Cléopâtre; liquid tree

Fully AI-Generated Movies in 2 years?; see these crazy Ai-made video ads

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 🔮.

Je m'appelle Thomas, plus d'infos sur moi en bas d'email.

Voici donc ma dernière sélection !

L’apéro

• Selfie de Cléopâtre et sa bande imaginé par MidJourney (source)

  

  • Voir aussi selfies de "Neanderthals during Stone Age" et "Napoleon and troops at Waterloo (1815)"

• Study Finds ChatGPT Outperforms Physicians (médecins) in High-Quality, Empathetic Answers to Patient Questions (source)

• Utiliser un ordi avec cette "souris" dans votre bouche controlée par votre langue ! (source)

  

  • the MouthPad: "Virtually invisible to the world, but always available to you, it is positioned across the roof of your mouth to put all of the power of a conventional touchpad in the tip of your tongue,"

  • The trackpad is tongue-sensitive, with right-clicks and left-clicks performed by a suck-in gesture and a tongue press, respectively.

  •  specifically designed to have a minimal impact on a user's speech, an important feature for anyone needing to speak to those around them

• Impressive 1-min video of a character completely created with 4 AI services 

  • Character Design: MidJourney ; Message: OpenAI ; Voice: ElevenLabs ; Video: @D-ID

• NASA managed for the first time to extract oxygen from simulated moon soil (NASA)

• Apple's cofounder Steve Wozniak: If You Want to Learn About AI Killing People, "Get a Tesla" (source)

• Demis Hassabis, previously CEO of Deepmind, is now CEO of Google Deepmind, the combination of Alphabet's 2 AI labs Deepmind and Google Brain

  • this means that the heart of Google’s AI operation shifts to London

  • London-based DeepMind was considered the company’s crown jewel, largely closed off from the rest of the group.

  •  Its mission to “solve” intelligence was sacred. Success for its employees was measured through publishing work in top-tier scientific journals like Nature. Three people close to DeepMind said its leaders are focused on “Nobel-level problems” or problems that, if solved, would be worthy of a Nobel Prize.

  • But now, “Demis is very pragmatic, and highly, highly competitive,” one former DeepMind employee said. “He is ideological in terms of the . . . vision, but he wants to be on the frontier of this. He’s not anti-commercial. This is the best way for him to do the job, or he wouldn’t have agreed to it.”

• AI vocabulary illustrated with one example (source):

  • With just a Large Language Model (LLM), we can look up the best restaurants in a given city.

  • With an agent, we can tell it to look up the highest rated restaurant with a table available and book the table for two.

  • With an autonomous agent, we can ask it to find the best restaurant that fits into my schedule and my preferences then book it for me and my best friend. 

    • Autonomous agents can do this by breaking down a task into subtasks and using memory between each step to guide the agent’s actions.

• An AI learnt how to play virtual tennis...after watching enough tennis videos (video, source)

  • Alors que Yann LeCun, directeur de la recherche chez Meta disait il y a 2 semaines : "S’entraîner sur du texte, c’est « facile » : il suffit de prédire un score de probabilité pour chacun des mots du dictionnaire. Et ça, aujourd’hui, on ne sait pas le faire avec de la vidéo. Voilà un grand défi pour les prochaines années : comment faire en sorte que les machines apprennent par l’observation de vidéo, d’images en mouvement ? »

• And, really cool, robots learnt how to play football in the real world (using AI the virtual world first of course): make sure to watch the first seconds of that video (more videos here)

  

• We generally think of autonomous driving as deep into a new winter, but Cruise (acquired in 2022 by General Motors) just extended its robot taxi service to run across all of San Francisco. (source)

• GPT-4 can do more causal reasoning than you think : it achieved new state of the art on a wide range of causal tasks (source)

  • “LLMs bring capabilities so far understood to be restricted to humans, such as using collected knowledge to generate causal graphs or identifying background causal context from natural language.”

• Top 100 of AI-generated music

  • See also Apollo - a streaming platform just for generative music.

• Really cool 50-sec video: Artist paints in virtual reality, adopting the term ‘Volumism’ to describe her 3D immersive creations of sculptured paint

• This AI-generated beer commercial looks exactly like how an alien intelligence would understand our beer commercials (watch the weird 30-sec video)

  • et dans le même genre, watch this crazy 30-sec AI Generated Pizza Commercial

• Japanese city asked its staff to experiment with ChatGPT for administrative tasks (source)

• Marvel Director Expects Fully AI-Generated Movies Within Two Years

  • "You could walk into your house and tell your AI: Hey, I want a movie starring my photoreal avatar and Marilyn Monroe's photoreal avatar," Russo explained. "I want it to be a rom-com because I've had a rough day, and it renders a very competent story with dialogue that mimics your voice." (source)

  • ça cadre bien avec cette tribune que j'ai écrite en 2017 pour le Journal du Net "Demain l’IA saura générer un épisode de Game of Thrones en une seconde"

• Mark Zuckerberg says Meta wants to ‘introduce AI agents to billions of people’ (source)

  • “We’re exploring chat experiences in WhatsApp and Messenger, visual creation tools for posts in Facebook and Instagram and ads, and over time video and multi-modal experiences as well,” Zuckerberg said

  • While Meta released an AI language model called LLaMA to researchers earlier this year, it has yet to debut anything resembling ChatGPT in a way that is widely accessible.

• IBM to Pause Hiring for Jobs That AI Could Do (Bloomberg)

  • "I could easily see 30% of these jobs getting replaced by AI and automation over a five-year period," said its CEO, which means that in total, AI could replace up to 7,800 jobs.

• Scientists noticed a surge in activity in an area of the brain associated with consciousness in dying patients even after their hearts stopped beating (source)

  • findings that could help explain why many report vivid visions during near-death experiences.

• Example of technological abandonment: the decline of mechanical car washes in the UK.

  • From 2000 to 2015, the number of them installed in roadside garages fell by more than half (from 9,000 to 4,200). By 2015, the vast majority of car washes in the country are done by hand. 

  • Why did automation in the car-washing world go into reverse gear? The Car Wash Association blames immigration, among other factors. In 2004, ten Eastern European countries joined the European Union, and migrants from these countries who arrived in the UK worked for such low wages that they were able to undercut the price of more productive—but also more expensive—mechanical car washes. In this case, cheaper human beings actually managed to displace the machines. (from the book A World Without Work)

• Scientists questioned whether the space tourism industry has even begun grappling with the concept of customers potentially having sex and conceiving children in space, which is a big deal given the growing body of evidence suggesting that microgravity and radiation could seriously mess up a fetus. (source)

• China plans to complete the first stage of its "International Lunar Research Station" on the Moon should by 2030 (source)

  • “Lunar soil will be our raw material and it will be 3D-printed into construction units.”

• Fighting CO2 pollution in cities with... "Liquid trees": A device with microalgae which binds to  carbon dioxide and produce pure oxygen through photosynthesis. (source, see video)

  

  • The microalgae replace two 10-year-old trees or 200 square meters of lawn.

  • the advantage of microalgae is that it is 10 to 50 times more efficient than trees. 

  • The team behind LIQUID 3 has stated that their goal is not to replace forests or tree planting plans but to use this system to fill those urban pockets where there is no space for planting trees. 

  • In conditions of intense pollution, such as Belgrade, many trees cannot survive, while algae do not have a problem with the great levels of pollution.

• New well-financed startup by ex-Apple employee just launched a screenless device that uses a combination of voice and gestures for input and can display information by projecting it onto nearby objects. (source)

  

Partagez cette newsletter par Whatsapp en cliquant ici ❤️

• Si vous appréciez cette synthèse gratuite, n’hésitez pas à prendre 3 secondes svp pour 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

À table !

🤨 A Spanish startup has built a sperm-injecting robot that can be controlled using a PlayStation controller. The team successfully used it to fertilize human eggs, eventually resulting in the birth of two healthy babies. (source)

• It's a robotic update to traditional in-vitro fertilization (IVF), which involves human specialists joining a woman's egg and a man's sperm in a dish using a special needle under a microscope, which sometimes — but not always — results in fertilization.

• Because the current IVF practice is so delicate and labor-intensive, it's very expensive, with each attempt at pregnancy in the US costing around $20,000, according to the report.

• The startup aims to make the process much cheaper and more accessible by increasingly automating parts of the process.

First machine to read someone's mind without brain implants (New York Times)

• Using a combination of fMRI scans and GPT tech, researchers have shown patients a film, and managed to decode their brainwaves to get the general idea of what the film is about.

  • Functional magnetic resonance imaging (fMRI) measures the small changes in blood flow that occur with brain activity.

• The model was trained to associate fMRI scan activity with known video/audio content, and then use that understanding to look at an fMRI scan and understand what the people were watching or listening to.

• It was also able to turn a person’s imagined speech into actual speech 

• and, when subjects were shown silent films, the system could "read" the person's brainwaves and generate relatively accurate descriptions of what was happening onscreen.

• using A.I. to translate the participant’s fMRI images into words and phrases.

• “We’re getting at meaning, something about the idea of what’s happening. And the fact that that’s possible is very exciting.”

• Example : While under the fMRI scan, the participants were also asked to silently imagine telling a story; afterward, they repeated the story aloud, for reference. Here, too, the decoding model captured the gist of the unspoken version.

  • Participant’s version: “Look for a message from my wife saying that she had changed her mind and that she was coming back.”

  • Decoded version: “To see her for some reason I thought she would come to me and say she misses me.”

• Limitations :

  • fMRI scanners are bulky and expensive. 

  • Moreover, training the model is a long, tedious process, and to be effective it must be done on individuals. 

    • When the researchers tried to use a decoder trained on one person to read the brain activity of another, it failed, suggesting that every brain has unique ways of representing meaning.

  • Participants were also able to shield their internal monologues, throwing off the decoder by thinking of other things. 

• And another team  was able to reconstruct visual images from fMRI scan data using text-to-image Ai Stable Diffusion. (source)

  • first row is the image presented to the test subject, second row is the reconstructed image from fMRI data.

    

Incroyable !!! Parrots learn to make video calls to chat with other parrots, then develop friendships (source)

• Scientists showed a group of parrots across a range of species and their volunteer caregivers to use tablets and smartphones how to video-call one another on Facebook Messenger.

• birds were first taught to ring a bell in order to signal that they wished to make a call

• Once the parrots rang their bells, caretakers presented them with a tablet home screen with pictures of possible friends to call, with pairs and trios of parrots grouped together mainly according to size and time zone.

• Once parrots learned to use the tablet interface, they were not rewarded with treats for doing so.

• During coordinated three-hour sessions, using their beaks to tap the screen, each bird could initiate up to two calls lasting no longer than five minutes each.

• They wondered: If given the choice, would the birds call each other?

• The answer was a resounding yes. “Some strong social dynamics started appearing,”

• Not only did the birds initiate calls freely and seem to understand that a real fellow parrot was on the other end, but caretakers overwhelmingly reported the calls as positive experiences for their parrots. 

• Some caregivers watched their birds learn skills from their video friends, including foraging, new vocalizations and even flying. “She came alive during the calls,” reported one.

• They formed strong preferences: a Goffin’s cockatoo, became fast friends with an African grey parrot “It’s been over a year and they still talk,” 

• The most popular parrots were also the ones who initiated the most calls, suggesting a reciprocal dynamic similar to human socialization. And while, in large part, the birds seemed to enjoy the activity itself, the human participants played a big part in that. 

• Some parrots relished the extra attention they were getting from their humans, while others formed attachments for the humans on the other side of the screen.

• Two sickly, elderly male macaws paired together in the study had scarcely seen another macaw in their lives, yet formed a deep bond—dancing and singing enthusiastically together through the screen and calling “Hi! Come here! Hello!” whenever one or the other moved out of the video frame.

• The findings suggest that video calls can improve a pet parrot’s quality of life.

• Example incroyabe des capacités vocales d'un perroquet : vidéo d'une minute

• Parrots' intelligence is extraordinary; certain species, like cockatoos and African greys, have demonstrated cognitive capabilities equal to that of an early-elementary-aged child.

  • The parrot brain has more than 3 billion neurons. "their total numbers of neurons are comparable to those of small monkeys"

  • 1 cm3 de cerveau de perroquet a deux fois plus de neurones qu'1cm3 de cerveau de primate (source)

  • Humans have 16 billion neurons in their cortex (and 85 billion in total in the brain) compared with 9 billion in gorillas and orangutans and 6-7 billion in chimpanzees.

---

Mes derniers podcasts (cherchez Parlons Futur dans votre appli de podcast préférée)

• Approximations et omissions dans le dernier C dans l’Air sur l’IA (Google, Apple, autre)

  • Plein de ressources utiles abordées pendant l'épisode à retrouver dans sa description

• Après ChatGPT, dites bonjour aux Action Engines (Google, Apple, Podbean)

• Après le texte et ChatGPT, la vidéo fera passer un nouveau cap à l’IA (Google, Apple, Podbean)

• Une super IA signifie-t-elle la fin de l’humanité ? Résumé de la thèse d’Eliezer Yudkowsky  (Google, Apple, Podbean)

• Et toujours : un entretien avec Jacques Attali

---

Les dernières newsletters :

• Starship : success or not?; Crazy story from OpenAI early days; Smart glasses tell you what to say on dates

• Mouse has 2 biological dads; GPT4 saved this dog; Bing writes winning caption for New Yorker's cartoon; AI creates 3D scene from 1 pic

• Superhuman: see how much work a professor did in 30 minutes with AI, + more gems about the future

• Best of GPT-4: all you need to know & gems you may have missed; See what the latest text-to-image AI can do (impressive)

• See what happened to rat after human mini-brain implant ; why China didn’t invent ChatGPT; crazy Biden audio deepfake

• Les précédentes

---

L’addition ?

Cette newsletter est gratuite, si vous souhaitez m'encourager à continuer ce modeste travail de curation et de synthèse, vous pouvez prendre quelques secondes pour :

• transférer cet email à un(e) ami(e) ou partager par whatsapp

• étoiler cet email dans votre boîte mail

• cliquer sur le coeur en bas d’email

Un grand merci d'avance ! 🙏

Ici pour s’inscrire et recevoir les prochains emails si on vous a transféré celui-ci.

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, ainsi que de Yelda, “A voice assistant to answer every call” (notre itw sur BFM Business “À la mairie de Plaisir, c'est l'intelligence artificielle qui répond au téléphone”), une des Future 40 startups de Station F.

• Je suis basé à Singapour (mon Linkedin, mon Twitter), également membre du think tank NXU.

Merci, et bon weekend !

Thomas