🚀 Crazy rocket video; sarco suicide pod; Zuck: "this device is the next foundational shift in personal computing"; fastest brick-laying robot
New Chinese Nuclear Submarine Sinks; Altman: "superintelligence in a few thousand days possibly"
Bonjour,
Vous recevez 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 🌏 et défense ⚔️ pour mieux appréhender le futur 🔮.
Je m'appelle Thomas, co-fondateur de YeldaAI, qui développe des IA pour répondre au téléphone en langage humain pour les administrations et les entreprises. Plus d'infos sur moi en cliquant ici.
Et voici donc ma toute dernière sélection !
L’apéro
Les derniers épisodes du podcast Parlons Futur (Spotify, Apple)
A graphic novel made (almost) entirely in Stable Diffusion. (source)
A surreal, delusional generative AI video - must-see (video)
Fuite d'un document interne révélant "les secrets" de MrBeast, le youtuber le plus suivi au monde (le document, extraits dans l'ADN)
including how to onboard new hires, the kinds of metrics that he thinks about to succeed on YouTube, etc.
"This is some of the most incredible rocket footage I’ve ever seen." Watch the 50-sec video
A Chinese rocket narrowly missed a landing on Sunday—the video is amazing
Deep Blue Aerospace is just one of several Chinese companies working on vertical landing.
La compète s'intensifie !
Wow: NASA Activates Ancient Thrusters on Voyager 1, the Most Distant Human Object in Existence It's incredible that it still has fuel left (source)
sonde lancée en 1977, il y a 47 ans !! Aujourd'hui distante de la Terre de bientôt 250 milliards de km (presque 20 fois la distance Terre-Soleil !)
Xavier Niel rejoint le board de TikTok (où se trouve à ce jour le founder chinois, un VC chinois et 2 financier américains) (Wired)
Sam Altman: "It is possible that we will have superintelligence in a few thousand days (!); it may take longer, but I’m confident we’ll get there." (source : voir plus bas partie abonnés)
Sam Altman reportedly poised to get equity in OpenAI for the first time (Techcrunch)
Follows the abrupt departure of OpenAI’s CTO, Mira Murati, and plans to restructure into a for-profit benefit corporation — similar to rivals such as Anthropic and Elon Musk’s xAI.
This is intended to make OpenAI more attractive to outside investors, who’ve chafed at the company’s current cap on returns.
But it’s likely to raise concerns from the AI safety community about whether OpenAI can hold itself accountable in its pursuit of superintelligent AI.
OpenAI bloque le Voice Mode de ChatGPT en Europe (source)
Under a strict reading of the AI Act, ChatGPT advanced voice is *illegal* in EU workplaces and schools because the system can recognize a user’s emotions. That’s prohibited by the AI Act.
OpenAI Just Released Its Long-Awaited "Strawberry" Model (The Verge)
It "performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology."
does a better job at writing code and solving multistep problems than previous models. But it’s also more expensive and slower to use than GPT-4o
doesn't yet have many of the features that make ChatGPT useful, like browsing the web for information and uploading files and images,"
The company claims it could be used "by healthcare researchers to annotate cell sequencing data, by physicists to generate complicated mathematical formulas needed for quantum optics, and by developers in all fields to build and execute multi-step workflows."
But as early testers have already discovered firsthand, it's still miles away from replacing a human scientist or coder. (source)
In fact, if recent posts making their rounds on social media are anything to go by, the o1-preview is still often struggling with the absolute basics.
What people do with ChatGPT (source)
Analysis of a million ChatGPT interaction logs
Great! Chatbots can persuade people to stop believing in conspiracy theories (MIT Tech Review)
Researchers from MIT Sloan and Cornell University found that chatting about a conspiracy theory with a large language model (LLM) reduced people’s belief in it by about 20%—even among participants who claimed that their beliefs were important to their identity.
Apparently, doctors in the US are now using LLMs to argue with insurers. (NYT)
This is pretty much exactly the joke from last year that half of LLMs will be turning three bullet points into three paragraphs and the other half will be summarising three paragraphs into three bullet points, says tech analyst Ben Evans
And in the UK, One in five doctors use AI such as ChatGPT for daily tasks, survey finds (Guardian)
Doctors are using the technology for activities such as suggesting diagnoses and writing letters
En France, le docteur Philippe Juvin illustre la charge administrative délirante des médecins avec une anecdote
Il doit scroller sans fin pendant plusieurs minutes dans une liste de centaines d'options pour caractériser la chute d'un patient !!!
Jensen Huang, founder & CEO of Nvidia said that coding is going to be mostly AI...
...and recommends to young people that they consider careers in biology, manufacturing and farming.
The best AI feature Google just announced: smart screenshots. (The Verge)
Pixel Screenshots uses AI to extract information from the screenshots you take, allowing you to search through them.
Electronic Arts demonstrates in a concept video a system that allows users to create a video game entirely from scratch
In the concept video, two gamers off-screen create a Minecraft-esque video game involving a maze of cardboard boxes.
Through a series of quick prompts, the users make the maze more complex while adding rules such as that characters can only die if they are hit with a grenade.
The changes show up pretty much instantaneously, and with zero technical ability required, and allow the game to be played immediately.
Meta has scrapped celebrity AI chatbots after they fell flat with users (source)
Turns out nobody wanted to chat with a random AI celebrity after all!…
…maybe people prefer to pretend they're celebrities and have millions of AI-generated fans to comment on their posts??
Well, that's coincidentally the weird idea behind the startup project SocialAI (not run by Meta) (Wired)
There’s only one real human in the SocialAI equation. That person is you.
When you first sign up, you’re prompted to choose these AI character archetypes: Do you want to hear from Fans? Trolls? Skeptics? Odd-balls? Doomers? Visionaries? Nerds? Drama Queens? Liberals? Conservatives?
You post text and then, instantaneously, several comments appear, cascading below your post, each and every one of them written by an AI character, these AIs also talk to each other.
"Just make a post, and watch as you get up to thousands, even millions, of AI-generated comments posing as humans in return, like you'd just went viral."
South Korean high schools have a big porn deep-fakes problem (BBC)
China Embarrassed When Brand-New Secret Nuclear Submarine Instantly Sinks. (WSJ)
When an electric truck crashes and catches fire, you need a LOT of water. Possibly even a plane full of water. (source)
Most electric-car batteries could soon be made by recycling old ones (The Economist)
Recent breakthroughs in recycling, together with a spate of technological improvements, mean that within a decade or so most of the global demand for raw materials to build new batteries could be met by recycling old ones.
Example: a new process using microwaves was able to recover 87% of the lithium in a used cathode in 15 minutes, vs 12 hours necessary to collect the same amount using a conventional method
Well, well, well: Elon Musk built a data centre for his AI startup, couldn’t get enough power hookups fast enough, so just bought and installed a bunch of (highly polluting) gas turbines without telling anyone, let alone getting a permit. (NPR)
Near-total collapse in venture-backed startups in China (FT)
The Financial Times has a big splash claiming a near-total collapse in venture-backed startups in China:
partly because US investors are pulling back
but much more because of the changed domestic political environment.
Apparently funding has gone from ~$100bn in 2022 to only $6bn in 2024 YTD (year to date), and 261 companies in 2024 versus a peak of 51k in 2018.
It’s very hard for governments to create startup ecosystems (as the EU has discovered) but it’s very easy to kill them, says tech analyst Ben Evans
“Orion glasses are the first device that is powered by our wrist-based neural interface,” Zuckerberg said. (Techcrunch, Wired)
a wrist-worn wearable prototype (not consumer-ready). Wearers can gesture while wearing the wristband-type peripheral to navigate around apps on the paired Orion glasses.
Meta’s Orion glasses — which remain very much a concept at this stage — are true AR.
Orion utilizes tiny projectors built into the glasses’ temples to create a heads-up display.
the images don’t appear as flat, 2D graphics in front of your eyes but that the virtual images now have shape and depth. “The big innovation with Orion is the field of view of 72 degrees" (most headsets are in the 30- to 50-degree range)
a futuristic pair of smart glasses that Zuckerberg hopes will lead the next foundational shift in personal computing.
Google paie 2,7 milliards de dollars pour faire revenir un génie de l'IA parti fâché en 2021 (source)
Noam Shazeer, a 48-year-old software engineer who was first hired by Google as one of its first few hundred employees back in 2000, left the company in 2021 after it refused his request to release a chat bot that he had developed with a colleague, Daniel De Freitas.
Google paid Character.AI $2.7 billion to license its technology as well as to get Shazeer and his team to agree to work for the company, according to The Wall Street Journal
Eric Schmidt, former CEO of Google, was reportedly impressed with Shazeer — so much so that he was convinced he would be able to build an AI model that could operate with human-level intelligence, according to the Journal.
MiniMax, l'IA chinoise pour générer des vidéos convaincantes à moins de 30 centimes (JDN)
L'éditeur d'IA chinois MiniMax a présenté un modèle de génération de vidéos. Les résultats sont convaincants et s'approchent de ceux de Sora.
Bien que le site soit en chinois simplifié, son utilisation est assez simple. Les prompts peuvent être rédigés en français.
Video-01 permet de générer des vidéos en format MP4 de 1280x720.
Point limitant majeur, les vidéos générées ne peuvent excéder les 6 secondes pour le moment.
AI put in charge of setting variable speed limits on US freeway (New Scientist)
Horrible ? : Sarco suicide pod: Arrests after American woman dies in Switzerland in controversial machine (CNN)
The so-called “Sarco” capsule had been deployed in a wood
Cast along sleek, aerodynamic lines, the “Sarco” causes death when its occupant releases nitrogen gas inside, lowering the amount of oxygen to lethal levels.
It is the brainchild of Philip Nitschke, an Australian physician famous for his work on assisted suicide since the 1990s.
Researchers have discovered a protein that, when deleted, could extend our lifespan and improve our quality of life as we age (new Atlas)
the IL-11 protein increases with age and spurs fat accumulation in the liver and abdomen, as well as reduced muscle mass and strength.
When that protein was suppressed in mice with a therapeutic, they found a 24.9% average increase in lifespan.
The mice also showed signs of improved metabolic function and increased calorie-burning brown fat, as well as a lower frequency of cardiometabolic diseases.
Altman’s $3.7 Billion Fusion Startup Leaves Scientists Puzzled (Bloomberg)
With significant backing from OpenAI’s billionaire CEO, Helion Energy promises to deliver a fusion power plant by 2028. Its results so far remain mysterious.
At the company’s headquarters, workers have their own reservations. In text messages and other communications reviewed by Bloomberg News and in interviews, current and former Helion employees expressed frustration with the company’s progress and skepticism that its seventh fusion prototype, named Polaris, will be operating to satisfy its internal year-end deadline.
“It doesn’t sound feasible,” one employee wrote in a message to colleagues. Another replied, saying that in private conversations, even CEO Kirtley didn’t sound certain.
See the World's fastest brick-laying construction robot in action: the one-minute video
Sapiens book by Yuval Noah Harari published in 2011 didn't age so well
He wrote then: Nationalism is “fast losing ground”; “there is at last real peace and not just the absence of war”; “the era when humankind stood helpless before natural epidemics is probably over”.
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 à vous abonner à la version payante pour recevoir lundi les résumés de ces longs articles le plus souvent issus des plus grands journaux (et accéder aux archives) en cliquant sur le bouton “Upgrade to paid/Subscribe now” plus bas :
Why Mark Zuckerberg thinks AR glasses will replace your phone
Why the end of Moore’s law will not slow the pace of change
Why language models won't create bioterrorists any more readily than celebrity cookbooks have created Michelin star chefs.
The Two Leading Theories as to why the Industrial Revolution started in Britain explained and summarized
Essay written by Sam Altman: The Intelligence Age
McDonald’s touchscreen kiosks were feared as job killers. Instead, something surprising happened
Les dernières newsletters :
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
et ici pour la version 100% en anglais
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.
Merci, et à bientôt !
Thomas