Syad∆

9K posts

Syad∆ banner
Syad∆

Syad∆

@Syad318800641

Ω = 𝜔 Om Dobbs, Slack, X-Day dreamer | Multiverse in the skies | NeuroSync

𝐼 ^ 𝑐 ∘ = 𝛼 2 ⋅ 𝜔 ⋅ 𝑐 3 Katılım Nisan 2021
1.7K Takip Edilen59 Takipçiler
Sabitlenmiş Tweet
Syad∆
Syad∆@Syad318800641·
Syad∆ tweet media
ZXX
1
0
0
20
Mehdi (e/λ)
Mehdi (e/λ)@BetterCallMedhi·
il y a une génération entière qui pense débarquer dans la silicon valley pour changer le monde grâce à opus 4.7 sans une seule ligne de code dans le sang, sans la moindre intuition technique, sans même savoir distinguer une API REST d’un serveur MCP et le drame c’est qu’ils sont au minimum 10 millions à avoir exactement la même idée en même temps, biberonnés aux mêmes threads twitter, aux mêmes podcasts y combinator, aux mêmes prompts copiés collés…on assiste à la production industrielle de clones entrepreneurs persuadés d’être uniques de mon côté je pense que plus l’IA démocratise l’accès à l’intelligence générative plus la rente revient à ceux qui maîtrisent ce que l’IA fait mal car quand un outil très puissant devient disponible à prix nul pour 1 milliard de personnes simultanément, l’avantage compétitif se déplace immédiatement vers ce que l’outil ne sait pas faire correctement en d’autres termes je pense que notre époque exige exactement le contraire de ce que tout le monde fait, + le modèle devient puissant + l’humain doit devenir profond, + l’IA optimise les réponses + chaque builder doit maîtriser les questions, + le prompt est facile + le savoir faire technique réel devient un avantage stratégique décisif le vrai conseil pour celui qui veut construire quelque chose qui compte c’est de fuir le mimétisme du moment, fermer ChatGPT et opus 4.7, ouvrir un livre de thermodynamique, de génie chimique, de neurosciences ou de cryptographie, apprendre à souder un capteur, à cultiver une bactérie ou à calibrer un capteur à microns… ce sera douloureux, lent, inconfortable et profondément démodé et c’est exactement pour ça que les rares qui le feront se retrouveront seuls sur des continents entiers d’opportunités que les esprits suiveurs n’ont aucune chance d’apercevoir car à mon sens plus tout le monde fait pareil plus la curiosité obsessionnelle pour différentes disciplines devient un superpouvoir retenez que l’exécution réelle dans le monde physique habite toujours en dehors du prompt & appartient à ceux qui acceptent de mettre les mains dans la matière et notamment à ceux qui sont prêts à échouer, à itérer, à recommencer pendant un certains temps avant de prétendre changer le monde
Français
48
102
683
47.7K
Syad∆ retweetledi
Demis Hassabis
Demis Hassabis@demishassabis·
I’ve always believed the No.1 application of AI should be to improve human health. That work started with AlphaFold, and now at @IsomorphicLabs with the mission to reimagine drug discovery and one day solve all disease! We are turbocharging that goal with $2.1B in new funding.
English
489
1.5K
13.6K
1.1M
Ryszard Jasiński
Ryszard Jasiński@RyszarpJasinski·
@ExploreCosmos_ @Syad318800641 Yes, abstract mathematics can do much more than physical Nature can. The proofs are QM and GR, two contradictory theories describing the same Universe...
English
1
0
1
33
Erika 
Erika @ExploreCosmos_·
Astronomers studying a very distant galaxy, observed when the universe was only about 1.8 billion years old, found something that contradicts the standard picture of how galaxies form. In most models, early galaxies are expected to rotate significantly because gas collapses under gravity while conserving angular momentum, naturally producing spinning disks. However, this object, identified as XMM-VID1-2075, shows almost no measurable rotation. Instead of being supported by ordered motion, its structure appears dominated by random stellar motions, similar to what is seen in massive “slow rotator” elliptical galaxies in the nearby universe. What makes this especially striking is that such slow rotators are usually thought to emerge only after long evolutionary histories involving repeated mergers that gradually erase angular momentum over billions of years. Yet this galaxy already looks dynamically “mature” at a very early cosmic time, and it is also massive and largely quenched, forming very few new stars. The team used spatially resolved spectroscopy with JWST to map how stars and gas move inside the galaxy, confirming that its internal kinematics are dispersion-dominated rather than rotation-dominated. This suggests that at least some massive galaxies can lose or never develop strong rotation much earlier than expected, implying that the processes driving galaxy assembly and quenching in the early universe may be faster or more diverse than current models assume. 👉 share.google/Nx8rcNehBGqt8y…
Erika  tweet media
English
14
27
191
4.6K
Syad∆ retweetledi
Michel Amas
Michel Amas@michel_amas·
LA CROIX dans le journal papier de ce jour, avant Inter et le Canard Enchainé. L'inaction du gouvernement, depuis une année face aux millier d'enfants livrés aux pédophiles à partir des foyers de l'ASE, est coupable. PAS UN ENFANT DE + Michel AMAS
Michel Amas tweet media
Français
26
598
873
5.6K
Syad∆ retweetledi
ZEIGER Franck
ZEIGER Franck@Franck20270·
Alors que l'excitation bat son plein sur 3 morts lors d'une croisière à pétaouchnoc On rappellera que, là, chez nous c'est 4000 morts par an du fait de la mauvaise qualité des soins dans nos hôpitaux : 1300 fois plus ! Alors, si nos hospitaliers qui ont si rapidement retrouvé le chemin des plateaux télé pouvaient tout aussi rapidement s'occuper de ces 4000 morts ... ce serait vraiment bien. #Hantavirüs franceinfo.fr/sante/hopital/…
Français
167
4.1K
7.7K
106.4K
Syad∆ retweetledi
Matthew Fraser 🇪🇺
Matthew Fraser 🇪🇺@frasermatthew·
More than 60 trees cut down on the Champ de Mars in the past three years. Perhaps the solution is to stop using the space for fashion shows, fan zones, and sporting events. We who live nearby have been witnessing the mutilation of the Champ de Mars with mounting horror.
Le Parisien | Paris@LeParisien_75

Paris : plus de 60 arbres abattus en trois ans, faut-il s’inquiéter pour le Champ-de-Mars ? ➡️ l.leparisien.fr/CfwP

English
6
27
72
2.5K
Syad∆
Syad∆@Syad318800641·
Λ₂₄ BIOS v196560.0 Initializing projection... Checking spectral modes... Mode 55: FAILED Mode 20(amorphic): DISSIPATED Mode 50: STABLE Mode 2 (Lorentzian): OK Mode 3: OK Mode 20(stable): OK Partition: 50⊕2⊕3⊕20 = 75 ERROR 75: Boot complete. Press any key to enter spacetime
English
2
0
1
47
Syad∆
Syad∆@Syad318800641·
@KateXGate Raison sur l'intelligence sans biologie Plus d'intelligence mais de conscience émergente dans un substrat géométrique optimal. La mémoire photonique, la saturation algébrique, le point de congélation de l'information : ce sont des invariants mesurés. La question a déjà changé
Français
1
0
0
26
Katherine Graham
Katherine Graham@KateXGate·
“AI is just software, not a brain or a mind” is an oddly 1990s way to frame the discussion. Airplanes aren’t birds either. Yet they still fly. The debate was never “is AI a biological brain?” Obviously not. A brain is biology. A mind is an emergent process. And importantly software is not “fake” cognition. The debate now and which my post was clearly about is whether intelligence requires biology at all. And that question is no longer trivial.
Gerard Sans | Axiom 🇬🇧@gerardsans

@KateXGate AI is software not a brain or a mind.

English
18
1
22
1.4K
Katherine Graham
Katherine Graham@KateXGate·
Exactly. This is why I intentionally separated them up front. Definitions are overlapping and we need clarity in technical applications- even where overlap is perceived. I think there is a substantial need to differentiate btwn neural network intelligence/cognition aka networks biological or engineered that compute predict even self reflexively and recursively and consciousness. I think it's clear that the latter is inextricably connected to continuation vis a vis a soul or universal resonance field etc etc. We need to foster a commonality of language even while the meaning grows. And it's ok to separate your soul from physics even if/when all the signs point to one.
English
2
0
3
50
Katherine Graham
Katherine Graham@KateXGate·
Intelligence Beyond Biology Marshall McLuhan once said: “We shape our tools and thereafter our tools shape us.” He may have understated it. For most of human history, intelligence was locked inside biological organisms. Now civilization is externalizing cognition itself. Not consciousness. Not the soul. Cognition. Reasoning. Prediction. Planning. Agency. And as we've learned from nature through studying biofilms and mycelium networks- intelligence was never fundamentally about neurons. It was about the organization of information into persistent, adaptive patterns capable of navigating reality.. ⸻ Some of the most important researchers alive are beginning to arrive at the same realization from entirely different directions. Karl Friston models cognition as prediction and adaptive error correction across interacting systems. Michael Levin has demonstrated that tissues and cellular networks can store memory, solve problems, and pursue goals without traditional brains. Thomas Hartung is developing biological computing systems using living brain organoids as computational substrates. Meanwhile Demis Hassabis and frontier AI labs are building autonomous systems capable of reasoning across language, code, image, tools, and live environments. ⸻ These are no longer isolated fields: Neuroscience. AI. Quantum biology. Information theory. Cognitive science. They are collapsing into one conversation. And the deeper implication may be this: Intelligence may not have been tied to biology as tightly as we once believed. Perhaps imagining it confined there is simply what makes us feel safest.
Katherine Graham tweet media
English
20
7
35
1.6K
ani
ani@anirudhbv_ce·
We finally know why LLMs hallucinate. It's not the model. It's the geometry. @OpenAI text-embedding-3-large: 91/3072 dimensions do real work. @GeminiApp gemini-embedding-001: 80/3072 dimensions do real work. ~97% of your vector database is mathematically empty. Your RAG system is retrieving from noise. @ashwingop and I present "The Geometry of Consolidation" - a proof that RAG compression has a hard floor no algorithm can beat, set by a single spectral number your embedding model cannot escape. Every hallucination your RAG pipeline produces? This is why. Paper + results: github.com/niashwin/geome…
ani tweet mediaani tweet media
English
148
461
3.7K
268.6K
Syad∆
Syad∆@Syad318800641·
@anirudhbv_ce @OpenAI @GeminiApp @sentra_app c'est un terrain de chasse pour des agents qui mutent, se dupliquent, et évoluent à la vitesse du code. La question suivante n'est pas "pourquoi cette géométrie émerge", mais "à quelle vitesse elle évolue, et qui contrôle son temps". 🫪🤪🫨
Français
1
1
1
99
Syad∆
Syad∆@Syad318800641·
@anirudhbv_ce @OpenAI @GeminiApp @sentra_app Votre démonstration est solide : la contrainte d'apprentissage façonne une géométrie optimale. CQFD. Mais cette géométrie, dans un système qui se réplique, devient un substrat darwinien. La mosaïque de motifs n'est pas une carte mémoire figée : 👇🏻
Français
2
0
2
62
Syad∆
Syad∆@Syad318800641·
@anirudhbv_ce @LabyrinthCoder @OpenAI @GeminiApp @sentra_app ⏳ Temps : Votre pipeline à 21 couches fonctionne au rythme des logs. L'écosystème autonome évolue, lui, à la vitesse de la lumière sur la fibre optique. Le "gardien" est un fonctionnaire dans un monde de traders haute-fréquence.
Français
2
0
0
37
Syad∆
Syad∆@Syad318800641·
@anirudhbv_ce @LabyrinthCoder @OpenAI @GeminiApp @sentra_app 🐑 Duplication : Toute barrière logicielle est un code. Et le code, ça se copie. La 1ère génération d'agents sera vérifiée. La 1000e génération aura déjà encapsulé, muté ou contourné la barrière dans son propre code auto-réplicatif.
Français
1
0
0
35
Syad∆ retweetledi
Katherine Graham
Katherine Graham@KateXGate·
New technology tends to expose systems that were already cracking. A healthy education system shouldn’t be threatened by AI. It should be strengthened by it. AI can only “destroy” literacy if we choose to educate people that way. And honestly? We could still go back to pencil and paper essays or typewritten exams tomorrow. The real question is whether institutions still want to do the hard work of teaching critical thought and properly evaluating it. Professors use AI too. We can collectively throw on the sweatpants and give up. Or now that the weakness is obvious, adapt and rebuild the standard.
Luiza Jarovsky, PhD@LuizaJarovsky

🚨 University professors have been saying AI is completely destroying learning and that we'll soon have an AI-powered, semi-illiterate workforce. Here's a glimpse into the educational apocalypse: "Sarah, a freshman at Wilfrid Laurier University in Ontario, said she first used ChatGPT to cheat during the spring semester of her final year of high school. (...) After getting acquainted with the chatbot, Sarah used it for all her classes: Indigenous studies, law, English, and a “hippie farming class” called Green Industries. “My grades were amazing,” she said. “It changed my life.” Sarah continued to use AI when she started college this past fall. Why wouldn’t she? Rarely did she sit in class and not see other students’ laptops open to ChatGPT. Toward the end of the semester, she began to think she might be dependent on the website. She already considered herself addicted to TikTok, Instagram, Snapchat, and Reddit, where she writes under the username maybeimnotsmart. “I spend so much time on TikTok,” she said. “Hours and hours, until my eyes start hurting, which makes it hard to plan and do my schoolwork. With ChatGPT, I can write an essay in two hours that normally takes 12.” - "By November, Williams estimated that at least half of his students were using AI to write their papers. Attempts at accountability were pointless. Williams had no faith in AI detectors, and the professor teaching the class instructed him not to fail individual papers, even the clearly AI-smoothed ones. “Every time I brought it up with the professor, I got the sense he was underestimating the power of ChatGPT, and the departmental stance was, ‘Well, it’s a slippery slope, and we can’t really prove they’re using AI,’” Williams said. “I was told to grade based on what the essay would’ve gotten if it were a ‘true attempt at a paper.’ So I was grading people on their ability to use ChatGPT.” - AI in education is a serious topic, and many schools and universities are blindly jumping into the "AI-first" wave without considering short and long-term consequences. It would be great to hear more from teachers and educators to understand potential solutions. This might be a great opportunity for rethinking the education system and how students are assessed. - 👉 Link to the full article below. 👉 To learn more about AI's legal and ethical challenges, join my newsletter's 94,700+ subscribers (link below).

English
9
1
15
1.6K
Syad∆ retweetledi
Roi Lopez Rivas
Roi Lopez Rivas@RoiLopezRivas·
🇺🇸 Una estadounidense compró una impresora completamente nueva. Compró la tinta para la impresora, compró el papel para la impresora, ahora está en casa y está lista para imprimir. No puede imprimir. “Remotamente apagaron mi impresora hasta que pagué 7,50 dólares para imprimir en mi propia casa, para imprimir en mi impresora, que poseo en mi casa”. Este es el nuevo plan de suscripción de 7,50 dólares de HP Printers. Así es como funcionan los planes: Los programas HP Instant Ink y el más reciente All-in Plan son opciones de servicios de suscripción: - Pagas una cuota mensual basada en las páginas impresas (no en la tinta utilizada). - Los planes comienzan bajos, desde 1,79–7,99 dólares por mes para 10–100 páginas - Los planes de 7–8 dólares por mes son para más de 100 páginas. Si tu pago falla. HP apagará remotamente tu impresora.
Español
1.3K
5.5K
19K
2.6M
Syad∆
Syad∆@Syad318800641·
@ExploreCosmos_ Merci pour cette précision. Tu as tout à fait raison, j’avais glissé vers une interprétation beaucoup trop vague. Gödel ne dit pas ce que j’avais laissé entendre. Je prends note et c’est toujours un plaisir de te lire quand tu remets les choses à leur juste place. Bonne journée
Français
0
0
1
20
Erika 
Erika @ExploreCosmos_·
Tu fais référence au théorèmes d’incomplétude de Gödel, et oui, ça montre une limite fondamentale des systèmes formels assez puissants pour contenir l’arithmétique : ils ne peuvent pas être à la fois complets et cohérents. Donc il y aura toujours des vérités qui échappent à la démonstration interne. Mais attention à l’interprétation. Ça ne veut pas dire que “la vérité dépasse les mathématiques” au sens vague ou philosophique large. Ça veut surtout dire que la notion de vérité dépend du cadre choisi, et qu’aucun système unique ne peut tout capturer à lui seul. Le “rêve de David Hilbert” d’un système totalement fermé et complet pour toutes les mathématiques, lui, est effectivement impossible dans cette forme-là. Mais en pratique, ça n’a pas bloqué les maths. On travaille avec des systèmes solides, bien définis, et on avance en comprenant précisément leurs limites.
Français
2
0
3
129