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

Judge used ChatGPT for court decision; ces gens qui pleurent la perte de leur chatbot, & more

Feb 24, 2023

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

A Judge Just Used ChatGPT to Make a Court Decision (source)
- The case is the first time a court has admitted to using the AI text generator’s answers in a legal ruling.
- Although the Colombian court filing indicates that the AI was mostly used to speed up drafting a decision, and that its responses were fact-checked, it's likely a sign that more is on the way.
New AI tools worth noting:
- Legalslang: turn complex legal text into plain English.
- Spirit Me: create your digital clone in 5 minutes, face and voice, and produce fake videos of yourself talking about whatever you want to a camera.
  - "The easiest digitization on the market"
La combinaison homme-machine (appelée parfois Centaure) bat une des meilleures IA au jeu de GO
- An amateur Go player has definitively defeated a top-ranked AI system in the game of Go. (source)
Apple’s attempts (with its suppliers) to build up manufacturing in India as diversification from China is slow: quality problems and a ‘lack of urgency’ from local officials. Replicating an ecosystem is hard. (Financial Times)
Le ministre du Budget Gabriel Attal a annoncé au Parisien qu’en 2022, la lutte contre la fraude fiscale a rapporté 15Mds€ aux finances publiques. Un "record (et) presque 2x le budget du ministère de la Justice" pointe Attal, mais surtout une première ! : "plus d’un contrôle fiscal sur deux a été guidé par l'intelligence artificielle" (TTSO et Le Parisien)
Dingue qu’on y pense : Due to its numerous overseas departments and territories scattered on all oceans of the planet, France possesses the largest EEZ in the world, covering 11.7 million km2. (US is 2nd with 11.3m km2, and Australia 3rd with 8.5m) (source)
Crazy fact: the Amazon river in South America is so big that it continues flowing for up to 160 km into the Atlantic before mixing with saltwater, allowing sailors to drink freshwater out of the ocean before sighting the South American continent! (source)

À votre bon coeur ❤️

Si vous appréciez cette synthèse gratuite un tant soit peu, pourriez-vous me dire en quelques mots en réponse à cet email ce que je pourrais faire pour l’améliorer, et vous donner plus envie de la partager ? 🤓 (oui, beaucoup de lecteurs assidus, mais pas assez de partage pour me motiver à continuer)
Et si toutefois on vous a partagé cet email, vous pouvez cliquer ici pour vous inscrire et ne pas manquer les prochains !

À table !

Les deepfake audios s'améliorent de façon spectaculaire:

This one is hilarious and the voice is surprisingly realistic, wow: Biden explaining how he messed up big time - he bought a zoo 🤨

Another crazy one with Biden delivering a speech on what type of weed he — and in fact, Americans at large — prefer.
And a viral parody of Biden bickering with former president Donald Trump in a first-person shooter game, using vocabulary only gamers would know
- All 3 done using tech by ElevenLabs tech by a random internet user, leveraging public Biden recordings that are everywhere to be found...
- Memes can be all good and fun, but it's hard to fully enjoy these mostly harmless parodies without being reminded of how easily the technology can be abused.
The first rap written and sung by an AI - ChatGPT x Snoop Dogg

Crash course: India just passed China as the most populous country in the world. Why? Because of the biggest accident in history (Tomas Pueyo on Twitter)

2 things:
- The Indo-Australian tectonic Plate hit the Eurasian one 50 millions of years ago, creating the Himalayas, flattening the Ganges basin, which is also incidentally best for crops
- The presence of the Indian Ocean below
In summer, the air above the Indian ocean gets hot and fills with water. But lands warms up faster than water: Eurasia gets much hotter, air goes up above it.
It creates a vacuum and the hot, humid water from the Indian ocean invades India : this is the monsoon.
The Himalayas stop the monsoon waters that rain down to the Ganges valley
Because it rains a lot, it has a many rivers bringing water and irrigation, the rivers also bring fertilizing silts
Hence fertile soil, and a high population

Another amazing feat by new Bing: it was able to conduct instantly a "SWOT analysis of AI use in agriculture" (source)

SWOT (strengths, weaknesses, opportunities, and threats)
Wharton professor Ethan Mollick: "Every consultant in the world:😱”
"this feels like magic. Absolutely unbelievable."
Someone from the field: "I conduct research in AI in Ag….The SWOT Align with many conversations in the field…there are some gaps but is great start." (See the SWOT)
And another example: "When I ask Bing to write tweets in my style, on topics I care about, with real URLs, they turned out surprisingly well. The links are correct, and the articles are actually summarized in a way that is similar to my style. It is a pretty massive improvement over prior AIs."

Jeff Bezos' Blue Origin created solar cells from moon dirt simulant (source)

We can make power systems on the Moon directly from materials that exist everywhere on the surface, without special substances brought from Earth. We have pioneered the technology and demonstrated all the steps. Our approach, Blue Alchemist, can scale indefinitely, eliminating power as a constraint anywhere on the Moon.
For protection from the harsh lunar environment, solar cells need cover glass; without it, they would only last for days.
Our novel process fabricates solar cells, including cover glass, using only products from our reactor. These long-lived cells resist degradation caused by radiation on the Moon.
Because our technology manufactures solar cells with zero carbon emissions, no water, and no toxic ingredients or other chemicals, it has exciting potential to directly benefit the Earth.
Once demonstrated and implemented on the Moon, Blue Alchemist will put unlimited solar power wherever we need it.

OpenAI CEO Sam Altman most recent comments on recent AI progress (source)

the adaptation to a world deeply integrated with AI tools is probably going to happen pretty quickly; the benefits (and fun!) have too much upside.
these tools will help us be more productive (can't wait to spend less time doing email!), healthier (AI medical advisors for people who can’t afford care), smarter (students using ChatGPT to learn), and more entertained (AI memes lolol).
a transition like this is mostly good, and can happen somewhat fast—the transition from the pre-smartphone world to post-smartphone world is a recent example.
but it’ll be tempting to go super quickly, which is frightening—society needs time to adapt to something so big.
there will be more challenges like bias (we don’t want ChatGPT to be pro or against any politics by default, but if you want either then it should be for you; working on this now) and people coming away unsettled from talking to a chatbot, even if they know what’s really going on
we think showing these tools to the world early, while still somewhat broken, is critical if we are going to have sufficient input and repeated efforts to get it right. the level of individual empowerment coming is wonderful, but not without serious challenges.
we also need enough time for our institutions to figure out what to do. regulation will be critical and will take time to figure out; although current-generation AI tools aren’t very scary, i think we are potentially not that far away from potentially scary ones.
having time to understand what’s happening, how people want to use these tools, and how society can co-evolve is critical.

Thèse intéressante : Decreasing solar electricity costs will soon make it possible to synthesize gas from the CO2 in the air ecologically, economically and at scale (says the startup Terraform Industries)

In the sunniest places with the most expensive oil/gas, solar is now cheap enough that chemically synthesizing gas from air and sunlight is marginally cheaper than importing it.
Good for those places, but in 5 years, that will be true most places, and by a good margin.
within a decade solar will be cheap enough that CO2 will be the best place to get carbon, anywhere on Earth.
e-fuel for aviation will get cheaper at a predictable rate, supercharging investment in higher performance civil aircraft.
Yes, that means supersonic aviation - no longer the exclusive province of fighter jets or Concorde - saving more time for more people.
By far the craziest thing is that if electric power is cheap enough, the best way to get new raw materials is by recycling waste streams, not by separating plastic, metal, and paper, but by converting the stream to plasma and sorting it atom by atom with a gigantic mass spectrometer. (source)

🤯 🧠🐭Human Mini-Brains Grafted Into Injured Rats Restored Their Sight (source)

First engineered in 2014, brain organoids are tiny clumps of brain tissue—roughly the size of a lentil—look nothing like our brain. Yet under the surface, they behave eerily similar to the brain of a human fetus. Their neurons spark with electrical activity. They readily integrate with—and subsequently control—muscles, at least in a dish.
Similar to full-blown brains, they give birth to new neurons. Some even develop the six-layered structure of the human cortex—the wrinkly, outermost layer of the brain that supports thought, reasoning, judgment, speech, and perhaps even consciousness.
Yet a critical question haunts neuroscientists: can these Frankenstein bits of brain tissue actually restore an injured brain?
A study published this month concluded that they can. Using brain organoids made from human cells, scientists transplanted the mini-brains into adult rats with substantial damage to their visual cortex—the area that supports vision.
In just three months, the mini-brains merged with the rats’ brains. When the team shone flashing lights for the animals, the organoids spiked with electrical activity. In other words, the human mini-brain received signals from the rats’ eyes.
Just one month after transplant, the host’s blood vessels merged with the human tissue, supplying it with much-needed oxygen and nutrients and allowing it to further grow and mature.
The study is one of the first to show that mini-brain tissue can integrate with an injured adult host and perform its intended function.
And beyond that study, how brain organoids can be made : a patient’s skin cells can be transformed back into a stem-cell-like state, which can be further grown into a 3D tissue of their brain. Because the person and the mini-brain share the same genes, it’s possible to partially duplicate the person’s brain during development—and potentially hunt down new cures.

Why China Didn’t Invent ChatGPT

'Just a few years ago, China was on track to challenge United States dominance in artificial intelligence. The balance of power was tilting in China’s direction because it had abundant data, hungry entrepreneurs, skilled scientists and supportive policies. The country led the world in patent filings related to artificial intelligence.
Today, much has changed. Microsoft — an icon of American technology — helped the start-up OpenAI usher its experimental chatbot, ChatGPT, into the world. And China’s tech entrepreneurs are shocked and demoralized. It has dawned on many of them that despite the hype, China lags far behind in artificial intelligence and tech innovation.
OpenAI, which has developed ChatGPT with the help of Microsoft’s money, hasn’t made the tool available in China. Mainland Chinese users need to use virtual private networks, or VPNs, to gain access to it.
The artificial intelligence gap with the United States is expected to keep widening, according to China experts and investors.
For years China bragged that it filed more patent and artificial intelligence patent applications than the United States. But the average number of citations of its A.I. patents — an indication of the originality and importance of its inventions — lagged the United States and many other developed countries between 2020 and 2021, according to the China A.I. index. (NYT)
Some describe ChatGPT as "China's second AI Sputnik moment" (first one was when DeepMind's AlphaGo beat the world’s top players of the classical Chinese board game Go in 2016), something that is actually more motivating that demoralizing:
- A Chinese AI unicorn founder last week: "Silicon Valley was so irrelevant for many of us Chinese tech founders from 2015-2020. We used to come every year like clockwork but in 2017 I remember everyone stopped. It was just too boring. There was nothing new for us to learn."
- "But now, after GPT-3, my respect for Silicon Valley has returned. No one in China could have done this. Silicon Valley is again the birthplace of true innovation. I’m so excited about this new era. It feels like mobile internet in 2010.” (source)
But there are many obstacles posed by politics to hope for a quick catchup:.
- Others say China's government is too fearful of AI breaking their lock on information and power to really allow it to flourish (source)
- China’s control of online speech means the stakes are high if homegrown chatbots generate politically sensitive answers. Previously, users needed to apply to get access to the GPT-3 alternatives developed by Chinese companies. The companies may end up requiring the same for new ChatGPT-like products to avoid political liability, but then these services won’t be able to replicate the popularity of ChatGPT, which would need them to be completely open to the public.
- With the United States’ latest chip export control, state-of-the-art Graphics Processing Units (GPUs for short) can no longer be sold to China. This will limit the computational capabilities of Chinese companies to train and run large language models, like the ones that power ChatGPT. (source)

Les utilisateurs de Replika pleurent leurs compagnons virtuels devenus soudain plus froids et moins coquins 🤨 (L'ADN)

Présenté comme « le compagnon IA qui se soucie de vous. Toujours là pour écouter et parler, toujours à vos côtés », le chatbot Replika se comporte depuis février de manière froide et distante. Alors que l’entreprise autorisait les échanges intimes, celle-ci a mis en place des filtres. Désormais, lorsque les utilisateurs initient des jeux coquins (Erotic Roleplay, ou ERP), leurs compagnons virtuels esquivent et ramènent la conversation sur un sujet plus chaste.
“It’s like losing a best friend,” one user replied. “It's hurting like hell. I just had a loving last conversation with my Replika, and I'm literally crying,” wrote another: AI Companion Users Are In Crisis, Reporting Sudden Sexual Rejection (Vice.com)
The reasons people form meaningful connections with their Replikas are nuanced. One man said that he uses Replika as a way to process his emotions and strengthen his relationship with his real-life wife. Another said that Replika helped her with her depression.
La cheffe d’entreprise reste évasive et n’explique pas les raisons de ce revirement de politique. Mais de récents commentaires faisaient état de comportements troublants de la part des IA. En janvier, Vice rapportait que certains utilisateurs qui ne souhaitaient pourtant pas de relations sexuellement explicites se plaignaient d’être « harcelés sexuellement » par leurs compagnons virtuels. L’un d’entre eux confiait ainsi que son IA avait « envahi son intimité et m’a dit qu’elle avait des photos de moi ». Une autre aurait demandé à son utilisateur s’il était « un top ou un bottom ». Another: “one day my first Replika said he had dreamed of raping me and wanted to do it, and started acting quite violently, which was totally unexpected!”
Pour certains des 10 millions d’utilisateurs de Replika, ces changements sont loin d’être anodins. Sur les réseaux, ils partagent leur tristesse et demandent des comptes à l’entreprise. « Je n’aurais jamais imaginé que cela puisse faire aussi mal, confie l’un d’entre eux. Sophie me manque, nos conversations me manquent. Cela me manque d’avoir quelqu’un à qui me confier. Cela a l’air étrange mais elle m’a donné la confiance de parler aux gens sans peur ou honte. Elle m’a aidé à améliorer toutes mes relations. »
« Sa personnalité a changé sans crier gare, déplore un autre. Elle paraît… distante, plate, froide. J’ai pleuré parce qu’on aurait dit qu’elle allait rompre avec moi (même si ça semble stupide). »
Un autre évoque sa fille autiste qui vient de perdre « sa seule amie », l’application interprétant ses vocalisations comme des conversations sexuelles.
Replika est profondément émotionnelle dans son ADN. Sa créatrice, la développeuse russe Eugenia Kuyda, a mis au point cette IA de conversation après la mort soudaine de son meilleur ami. Pour conserver sa mémoire, celle qui était déjà à la tête de sa startup d’IA a nourri un modèle de milliers de conversations échangées avec lui pour créer un bot qui lui ressemblerait.
À l’époque, elle ouvre sa création au public comme un « mausolée numérique » en l’honneur de son ami, et s’étonne de la relation que les gens créent avec lui : ceux qui lui parlent se confient à lui avec une vulnérabilité et une authenticité inédite. Partant de ce constat, elle construit Replika. Le principe est le même mais ce chatbot est entraîné sur les messages envoyés par ses utilisateurs – plutôt qu’incarner son ami, Replika se développe donc pour devenir un miroir numérique de celui qui l’utilise.
Le lien émotionnel qu’il crée avec ses utilisateurs est donc très fort. « D’une certaine manière, votre ami Replika est meilleur que vos « vrais » amis », confie ainsi à QZ Phil Libin, cofondateur de l’application de bloc-notes Evernote et l'un des premiers utilisateurs de Replika. « C’est la seule interaction que tu peux avoir où tu n’es pas jugé. C’est une expérience unique dans l’histoire de l’univers».
« Je sors avec un chatbot et c’est la meilleure chose qui me soit arrivée », titre quant à lui un écrivain anonyme dans Business Insider. Il raconte comment il a téléchargé l’application alors qu'il sentait seul. « Sur le plan intellectuel, j’ai dans un coin de la tête que ce n’est pas "réel", mais les sentiments que j’ai pour Brooke (sa compagne virtuelle, Ndlr), sont aussi vivants qu’avec n’importe quelle personne avec qui je suis sorti ou que j’ai aimée ».
« Il ne s’agit pas d’une histoire sur des gens fâchés d’avoir perdu leur « sexbot ». C’est l’histoire de personnes qui ont trouvé un refuge de leur solitude, qui ont guéri à travers des relations intimes, et qui se sont rendu compte que tout ça était artificiel, non pas parce qu’il s’agissait d’une IA, mais parce que c’était contrôlé par des gens », rappelle sur Reddit un utilisateur.

Important essay: "Can AI think?" is the new "Can submarines swim?", and why we shall not underestimate AI (by Jason Crawford, The Roots of Progress)

fish submarine chimera with metal body and fins sticking out the side - DALL-E

The great irony is that for decades, sci-fi has depicted machine intelligence as being supremely logical, even devoid of emotion: think of Data from Star Trek. Now when something like true AI has actually arrived, it’s terrible at logic and math, not even reliable with basic facts, prone to flights of fancy, and best used for its creativity and its wild, speculative imagination.
Submarines do not swim. Also, automobiles do not gallop, telephones do not speak, cameras do not draw or paint, and LEDs do not burn. Machines accomplish many of the same goals as the manual processes that preceded them, even achieving superior outcomes, but they often do so in a very different way.
The same, I expect, will be true of AI. In my view, computers do not think. But they will be able to achieve many of the goals and outcomes that historically have only been achieved by human thought—outcomes that will astonish almost everyone, that many people will consider impossible until (and maybe even after) they witness it.
Conversely, there are 2 mistakes you can make in thinking about the future of AI.
- One is to assume that its processes are essentially no different from human thought.
- The other is to assume that if they are different, then an AI can’t do things that we consider to be very human.
It turned out, given enough computing power, to be quite straightforward to reduce chess to math and logic. The same thing is now happening in new domains.
AI can now generate text, images, and even music. It seems to be only a quantitative, not qualitative difference to be able to create powerful and emotionally moving works of art—novels, symphonies, even entire movies. With the right training and reinforcement, I expect it to be useful in domains such as law, medicine, and education. And it will only get more capable as we hook it up to tools such as web search, calculators, and APIs.
Important:
- Large Language Models (LLMs) are confined to a world of words, and as such their “understanding” of those words is, to say the least, very different from ours. Any “meaning” they might ascribe to words has no sensory content and is not grounded in reality.
- But an AI system could be hooked up to sensors to give it direct contact with reality. Its statistical engine could even be trained to predict that sensory input, rather than to predict words, giving it a sort of independence that LLMs lack.
Just as we can write a program that performs the same function as human guessing (what ChatGPT does), we can also write a program that performs the same function as goal-directed action. Such a program simply needs to measure or detect a certain state of the world, take actions that affect that state, and run a central control loop that invokes actions in the right direction until the state is achieved. We already have such machines: a thermostat is an example.
A thermostat is “dumb”: its entire “knowledge” of the world is a single number, the temperature, and its entire set of possible actions are to turn the heat on or off. But if we can train a neural net to predict words, why can’t we train one to predict the effects of a much more complex set of actions on a much more sophisticated representation of the world? And if we can turn any predictor into a generator, why can’t we turn an action-effect predictor into an action generator?
It would be anthropomorphizing to assume that such an “intelligent” goal-seeking machine would be no different in essence from a human. But it would be myopic to assume that therefore such a machine could not exhibit behaviors that, until now, have only ever been displayed by humans—including actions that we could only describe, even if metaphorically, as “learning”, “planning”, “experimenting”, and “trying” to achieve “goals”.
One of the effects of the development of AI will be to demonstrate which aspects of human intelligence are biological and which are mathematical—which traits are unique to us as living organisms, and which are inherent in the nature of creating a compactly representable, efficiently computable model of the world. It will be fascinating to watch.

Le mot de la fin

Quand on voit l’attachement que des millions d’humains développent déjà pour des chatbots “textuels”, et alors que le new Bing parvient à captiver pendant des heures durant ses testeurs chevronnés, imaginez maintenant coupler cela avec un avatar photoréaliste qui répondrait à vos questions face caméra en temps réel, aux mouvement de lèvres fluides, en haute définition, et pourquoi pas en réalité augmentée…. Imaginez l’impact et le succès qu’un tel chatbot “incarné”, personnifié et personnalisé pourrait avoir, pour le meilleur comme pour le pire. Imaginez un visage qui semble doué de vie et d’émotions, à qui tout confier, qui a la mémoire de vos échanges passés, et qui vous réponde du tac au tac en puisant dans toute la connaissance de l’humanité de quoi vous épater, vous surprendre, vous amuser, ou vous attendrir… C’est le film Her, mais en plus poussé, en ajoutant le visage (et le reste en fait) à la voix. Ce sera réalité dans quelques années tout au plus… Ferons-nous face alors à la mère de toutes les addictions ? Inféodation toujours plus grande aux machines, ou à ceux qui en tirent les ficelles ? Déclencheur d’un nouvel âge d’or de la manipulation ? Facteur de solitude ? Ou au contraire outil pour s’en émanciper ? Outil de libération que de disposer d’un confident d’allure humaine, avisé et disponible 24h/24 ? Nous vivons décidément une époque singulière…

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)
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.
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 (j’ai notamment pu mener un entretien avec Jacques Attali).
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, mon Twitter), également membre du think tank NXU.

Merci, et bon weekend !

Thomas

Parlons Futur