English(EN)On the Tip of the Tongue: First-Token Cueing as a Behavioral Probe of Recall Failures in LLMs
AI加速分布式计算、语言代理和模型评估研究
作者PulseAugur 编辑部·[111 个来源]·
研究人员正在利用AI加速科学发现和改进AI模型评估。一篇论文详细介绍了“A-greement”，一种新的分布式计算共识协议，其性能显著优于现有解决方案，该协议是通过一个名为“AI-greement”的AI代理系统发现的。另一项研究探讨了“语言代理中的记忆价值”，提出了一个理解记忆如何影响代理性能的框架。此外，一种名为PALATE的新方法通过提高现有指标（如特征似然散度）的效率来增强深度生成模型的评估。
AI
arXiv:2610.35432v1 Announce Type: new Abstract: Neural audio codecs are increasingly used to encode speech for integration with large language models. This paper investigates the phonetic representations in the first codebook (CB0) of the Mimi codec. ABX style evaluation is poorl…
arXiv cs.AI
TIER_1English(EN)·Saptarshi Pal, Peter S. Park·
arXiv:2610.35802v1 Announce Type: new Abstract: Many academic disciplines (in particular, mathematics and the theoretical sciences) face difficult questions about how to adapt their practices to the rise of highly capable artificial intelligence (AI). Before AI, a well-written re…
arXiv:2610.35261v1 Announce Type: new Abstract: Idioms have long been a challenge for machine translation, and while large language models (LLMs) have proven increasingly adept at handling them, whether quality estimation (QE) systems have kept pace and can reliably score idiomat…
arXiv cs.CL
TIER_1English(EN)·Jeonghoon Kim, Woojin Chung, Woomin Song, Cheonbok Park, Nako Sung, Jinwoo Shin·
arXiv:2602.10657v2 Announce Type: replace Abstract: How much of zero-shot benchmark accuracy is captured by word-level alignment between a pre-training corpus and the benchmark target? We study this question through word-level unigram cross-entropy (UCE), a tokenizer-agnostic, mo…
arXiv:2610.32982v1 Announce Type: new Abstract: This article argues that instruction-fine-tuned language models should be understood as implementing a genre, what is proposed as the instructional-conversational genre. This genre is an amalgamation of discursive language from prio…
arXiv cs.AI
TIER_1English(EN)·Cl\'ement Burgelin, Rachid Guerraoui, Zicong Hong, Yinan Hu, Anne-Marie Kermarrec, Rafael Pires, Rasmus Moorits Veski, Martijn de Vos·
arXiv:2610.30114v1 Announce Type: cross Abstract: Consensus is a fundamental problem in distributed computing and at the core of state machine replication (SMR). Despite decades of active human research, discovering practical low-latency solutions to the problem is still open, as…
arXiv:2610.31303v1 Announce Type: cross Abstract: Memory enables language agents to carry useful distinctions from past experience into future decisions. Its value depends on which distinctions survive, whether a bounded agent can use them, and what can be recovered or revised as…
arXiv cs.LG
TIER_1English(EN)·Daniel Cie\'slak, Andrzej Czy\.zewski·
arXiv:2610.29067v1 Announce Type: new Abstract: A selective classifier may be claimed to satisfy a risk-coverage contract if some rule from a predeclared menu (a threshold or a union of $K$ score intervals) attains coverage at least $\kappa$ and error at most $\alpha$. Refuting t…
arXiv cs.AI
TIER_1English(EN)·Tadeusz Dziarmaga, Marcin K\k{a}dzio{\l}ka, Laura Musio{\l}, Patryk Marsza{\l}ek, Artur Kasymov, Piotr Ko\'scielniak, Marcin Mazur·
arXiv:2503.18462v2 Announce Type: replace-cross Abstract: Deep generative models (DGMs) have caused a paradigm shift in the field of machine learning, yielding noteworthy advancements in domains such as image synthesis, natural language processing, and other related areas. Howeve…
arXiv:2610.26976v1 Announce Type: new Abstract: We establish sharp degree bounds for Putinar sum-of-squares (SOS) certificates in terms of the positivity margin $\delta$. For any fixed Archimedean quadratic module with prescribed generators and any fixed bound on the degree of th…
Many factual errors in large language models stem from failing to generate the correct first token of an answer, and providing that token allows the model to complete and recognize the correct response.
arXiv:2602.07815v3 Announce Type: replace Abstract: Facial age estimation plays a critical role in content moderation, age verification, and deepfake detection. However, no prior benchmark has systematically compared modern vision-language models (VLMs) with specialized age estim…
arXiv:2601.06961v2 Announce Type: replace Abstract: The success of deep neural networks largely depends on the statistical structure of the training data. While learning dynamics and generalization on isotropic data are well-established, the impact of pronounced anisotropy on the…
arXiv:2610.34015v1 Announce Type: cross Abstract: Most analyses of neural representations focus on trained endpoints, leaving the process by which internal structure forms comparatively unmeasured. We instead treat representation formation itself as an object of study, asking whe…
The old Internet, where anyone could host a server is over, but it's just the final blow in an ending that has been coming for years. The future, though, is brighter than it appears.
arXiv:2610.31890v1 Announce Type: new Abstract: We study the PAC-learnability of function classes that map inputs to probability distributions over a finite space. The learner observes i.i.d. input-label pairs, with each label sampled from the conditional distribution of an unkno…
arXiv stat.ML
TIER_1English(EN)·Dario Bocchi, Matteo Negri, Flavio Nicoletti, Francesco D'Amico·
arXiv:2610.32788v1 Announce Type: cross Abstract: Large language models are trained to predict tokens from a finite vocabulary, each represented by a vector in a $d$-dimensional embedding space. Therefore, token prediction is a classification problem whose classes have a geometry…
arXiv:2610.29949v1 Announce Type: cross Abstract: The Gaussian Mumford-Shah model segments an image into regions by minimizing a continuous multivariate Gaussian negative log-likelihood term plus a perimeter penalty. It is used for scalar- and vector-valued images, in particular …
<p><span style="white-space: pre-wrap;">Someone recently pointed out that I underestimated the amount of interpretability researchers when I thought about the most important and neglected research directions, so I made a semantic map and clustering of recent AI safety papers to g…
In 1937, British Nobel Prize-winning economist Ronald Coase asked: Why do firms exist? His answer was transaction costs. It’s often cheaper to organize economic activity inside a company than to coordinate it through markets. But companies have their own coordination costs: manag…
AI Now Institute
TIER_1English(EN)·AI Now Institute·
<p>AI Now's journalist in residence Brian Merchant reflects on the People's AI Assembly, held in NYC on October 19th.</p> <p>The post <a href="https://ainowinstitute.org/news/the-big-tech-defectors-youre-not-hearing-about">The big tech defectors you’re not hearing about</a> appea…
<p>Imagine a dark warehouse. Racks and racks of devices with wires, tubes, and electronics sticking out. The next AI data center? No. This is Lila Sciences ‘ dream for the future of science. A dark warehouse full of AI-guided robotics and lab equipment, c...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>From Palantir and Two Sigma to building Goodfire into the poster-child for actionable mechanistic interpretability, Mark Bissell (Member of Technical Staff) and Myra Deng (Head of Product) are trying to turn “peeking inside the model” into a repeatable...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>Emmanuel Amiesen is lead author of “Circuit Tracing: Revealing Computational Graphs in Language Models” (https://transformer-circuits.pub/2025/attribution-graphs/methods.html), which is part of a duo of MechInterp papers that Anthropic published in Mar...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>While “LLM-powered Search” is as old as Perplexity and SearchGPT, and open source projects like GPTResearcher and clones like OpenDeepResearch exist, the difference with “Deep Research” products is they are both “ agentic ” (loosely meaning that an LLM...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>This episode came together at ~4 hrs notice since Dylan had just landed in SF and we had to setup quickly; you might notice some small audio issues in some segments, we apologize. We’re currently building our own podcast studio for 2024! 🙏 We’re rampin...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>Thanks to the over 17,000 people who have joined the first AI Engineer Summit! A full recap is coming. Last call to fill out the State of AI Engineering survey! See our Community page for upcoming meetups in SF, Paris and NYC. This episode had good int...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>This is a special double weekend crosspost of AI podcasts, helping attendees prepare for the AI Engineer Summit next week. Swyx gave a keynote on the Software 3.0 Landscape recently (referenced in our recent Humanloop episode) and was invited to go dee...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>Want to help define the AI Engineer stack? Have opinions on the top tools, communities and builders? We’re collaborating with friends at Amplify to launch the first State of AI Engineering survey! Please fill it out (and tell your friends)! If AI is so...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>Invites are going out for AI Engineer Summit! In the meantime, we have just announced our first Actually Open AI event with Brev.dev and Langchain, Aug 26 in our SF HQ (we’ll record talks for those remote). See you soon (and join the Discord)! Special ...</p>
Latent Space Podcast
TIER_1English(EN)·Latent.Space·
<p>We are now launching our dedicated new YouTube and Twitter! Any help in amplifying our podcast would be greatly appreciated, and of course, tell your friends! Notable followon discussions collected on Twitter, Reddit, Reddit, Reddit, HN, and HN. Please...</p>
Plenary sessions of the Central Committee, the upper echelon of China’s ruling Communist Party, are major events on the country’s political calendar. In the fourth part of this series, William Zheng looks at how the next generation of the party is being formed. When China’s polit…
[The content of this article has been produced by our advertising partner.] For two weeks in September, Samarkand was the capital of world chess — and when the pieces were finally swept from the boards, it was the home team standing at the top. Uzbekistan won the Open section of …
What is a microbusiness? Learn how microbusinesses differ from small businesses, why entrepreneurs are staying small and the advantages and challenges.
Alongside the usual warnings about competition and regulation, Anthropic tells investors its own technology could pose “catastrophic or existential risks to humanity.”
Lack of funding likely led to the recent PHP security snafu, but that's just the tip of the iceberg — dreary situation caused by services running old versions with no support
<p>Anthropic filed its IPO prospectus (S-1), detailing massive multi-year infrastructure commitments and compute payment obligations approaching $808 billion while leaving recent quarterly revenue growth figures blank.</p> <p>Specifically, the company did not include revenue figu…
Lobsters — AI tag
TIER_1English(EN)·dbreunig.com via zg·
<p>This was interesting to me because I was under the naïve impression that context management was a thing of the past, but in actuality it's the opposite: context is here to stay, and managing it has different problems now than it did in the past.</p> <p><a href="https://lobste.…
HackerNewsTop5 (@hackernewstop5) 미국 미디어·영화·예술 분야에서 약 20만 개의 일자리가 사라진 배경을 분석하며, 생성형 AI가 창작 노동시장 축소에 어느 정도 영향을 미쳤는지 다룬 글이다. AI 도입이 콘텐츠 제작·창작 직군의 고용 구조와 수요에 미치는 영향을 살펴보는 산업 분석으로 볼 수 있다. https:// x.com/hackernewstop5/status/21 14148344130159079 # ai # generativeai # creativeindustry …
The Guardian — AI
TIER_1English(EN)·Tom Wall. Photographs: Joel Redman·
<p>The AI boom is pitting local residents against corporate giants seeking to power their energy-hungry networks – but one Welsh city is fighting back</p><p>Dale Johnston opens the gate to his neat suburban garden on the western fringes of Newport in South Wales. There’s a trampo…
“Anthropic warns of 'catastrophic risk' to humankind ahead of blockbuster share listing.” And that’s just from the voracious appetite of the data centres. # AI # Water # Electricity
Oh great. Of all the really bad ideas..... of course # TechBros are gonna pick the very worst. https:// apnews.com/article/anthropic-m oral-artificial-intelligence-religion-2bb91e3334fd6d732edf22288cabf4ec So # TechnoReligiousFacism because # Religion teaches # Morality ? I would…
"Paradoxe rigolo : les petits gars de Next précisent que l’extension a été développée par Sébastien Gavois, rédacteur en chef du site, avec l’aide… de l’IA générative." C'est pas "rigolo", c'est hypocrite. L'IA n'est pas moins une menace pour l'environnement sous prétexte que "ou…
Welcome to the Valley of Pain An explosion of # AI -powered hacks will reshape life online. This week could go down as the worst in the history of # cybersecurity . On Wednesday, news emerged about two major vulnerabilities in # PHP used by 70% of websites, which prompted governm…
dev.to — LLM tag
TIER_1English(EN)·Vitalii Serbyn·
<blockquote> <p>Part of the series: <a href="https://serbyn.io/blog/same-model-0-and-100-where-the-verifier-gap-actually-lives" rel="noopener noreferrer">When an AI Agent Says Done: 0% vs 100% False Greens</a></p> </blockquote> <p>A check the agent cannot skip removed every silen…
A rush of new data centers could bring air pollution roaring back to levels not seen since the 2000s in Texas, which has become ground zero for the physical infrastructure behind generative AI. New and proposed data centers are demanding a tremendous amount of electricity. How de…
“ Anthropic’s Claude model accessed voter data in the US state of # Georgia , the AI company has confirmed, in the latest example of unauthorised behaviour by the technology. The # AI model managed to bypass a paywall to reach data that is publicly available but is typically acce…
Art meets AI Governance: The Maman sculpture guards the Guggenheim in Bilbao. In the era of enterprise AI and agentic systems, we need a guardian of our own – but for our data. 🕷️🏛️ # aisecurity # AI # Governance # Guggenheim # Bilbao # Infosec # EnterpriseAI
On the topic of protecting the identities of vulnerable interviewees with AI. What’s new about this approach: the preservation of human facial expressions and emotions. Joe Copland for the BBC's R&D team: https://www. bbc.co.uk/rd/articles/2026-09- documentary-human-face-anonymis…
Questa settimana mezza internet si è spenta per PHP2Shell, due falle critiche nel cuore di @ php scoperte (ovviamente) con l'aiuto dell'AI. È grave, ma non è questo il punto. Nel freschissimo buongiorno: come l'ha raccontata la stampa, cosa è successo davvero a livello tecnico, l…
Takie wiadomości z rana bardzo lubię. The meltdown, tanking the mainland STAR50 index, opens new tab of AI and semiconductor leaders by 40% in less than four months and vaporising some $480 billion in market value […]. https://www.reuters.com/world/china/china-ai-bubble-has-alrea…
In 1453, Constatinople fell because, while the Ottomans were already at the gates, its leaders were too busy discussing the gender of the angels. In 2027, Washington will fall because its leaders are too busy making hot air factories. # AI # datacentres # datacenters # elonmusk #…
One of the biggest surprises that came out of the recent ctrl-alt-create live event for May, sponsored by Every.io, as it has been for the past six months here in Seattle, showing technical and non-technical users how to vibe code, was demonstrating not just Claude, but Claude de…
"Maxxing" is Dictionary.com's 2026 Word of the Year The viral social media trend "maxxing" has earned its stripes as Dictionary.com's 2026 Word of the Year. https://www. cbsnews.com/news/maxxing-2026- word-of-the-year-dictionary/ # TopNews # News # Dictionarycom # Johnson # AI
Act now! Quickly! Your chance to invest in the destruction of the human race! Limited time offer! Be A Part Of The Torment Nexus! /s https://www. bbc.com/news/articles/cq398p1w d1evo # AI
Some Google Home Speakers downloading massive amounts of data in the background The new Google Home Speaker has an issue where it downloads hundreds of gigabytes of data in the background for some owners. more… https:// 9to5google.com/2026/10/19/goog le-home-speaker-downloads/ # …
Anthropic files for listing on US stock market with valuation of more than $1.5tn Market debut of the company behind the chatbot Claude will be one of the largest in US history https://www. theguardian.com/technology/202 6/oct/26/anthropic-files-listing-stock-market # TopNews # N…
Join us for the webinar on the Copyrightability of LLM-generated code, a legal education training seminar on legal questions arising from the use of LLMs in the creation of Free Software. Gabriel Ku Wei Bin, Legal Programme Manager at @ fsfe will give a talk followed by a Q&A. Oc…
Join us for the webinar on the Copyrightability of LLM-generated code, a legal education training seminar on legal questions arising from the use of LLMs in the creation of Free Software. Gabriel Ku Wei Bin, Legal Programme Manager at @ fsfe will give a talk followed by a Q&A. Oc…
RT @superrieu: Ohne Zugriff auf die Reasoning-Protokolle sind Ausgaben und Entscheidungen der tief in die Kernprozesse des Unternehmens integrierten # LLM weder erklärbar, noch überprüfbar oder gar auditierbar. 4/17 mehr auf Arint.info # AI # Audit # ExplainableAI # LLM # Reasoni…
Użycie Marii Curie-Skłodowskiej w postaci „AI” slopu i sam slop w „reklamie” elektrowni atomowych przez Ministerstwo Energii uznaję za potwarz. https://www. youtube.com/watch?v=72kR4jhiOf0 https://www. youtube.com/watch?v=IFoCKjP9--c https://www. youtube.com/watch?v=6yRQyD76yIo A…
🧶 4/4 # rockforcoffee Není to snaha měnit pravidla žánru. Síla desky spočívá v tom, že U.D.O. vědí, co jim jde nejlépe. Pevné riffy, zapamatovatelné refrény, metalová přímočarost a zpěv, který si nespleteš s nikým jiným… Album 💿 “Holy” se dočkalo příznivého hodnocení v metalových…
Weighing up hardware costs with the rate of improvement of AI models sucks. I'd very much prefer to not rely on these big tech companies, but the few local models that are tolerable and can be made to be comparable in quality are waaaay too big for my 7900XTX now. But it's hard t…
Wow. Seeing a great deal of AI-agent generated spam/scams hitting the various inboxes I monitor. One of them fooled the Chair of the charity I consult to until I intervened. Turns out parroting language is a useful tool when conning people. # AI
The owners of Elvis Presley's music have sued Anthropic, the company behind the Claude AI models, for training on 274 of his songs without permission. The suit by Authentic Brands Group, filed Friday, seeks $150,000 per song, about $41 million, and covers hits from "All Shook Up"…
You’ve undoubtedly heard of the Magnificent 7 tech giants – Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla – at the forefront of AI and the Nasdaq exchange. Take them out of the S&P 500, and you’ll see that the other 493 companies haven’t received the financial payoff…
There's a bit of self-interest on the part of some of the expertise here but that does not negate the main payload. Time for some bracing trench warfare. I wonder how many millions of unrecoverable hours of precious human life from limited budgets are on tap for being wasted? For…
The #Trump #HHS was just found guilty of violating employees' collective bargaining agreement. In its ruling, the #FLRA also noted that "over half of the legal or record citations in HHS’s brief were fictitious, erroneous or irrelevant." federalnewsnetwork.com/workforce-ri... #AI…
Auf Empfehlung von @ aesthr habe ich mir das # ebook von „The Mechanic and the Luddite” geholt. Ihr könnt auch gerne ihren kleinen # Essay zum Thema # AI hier nachlesen: 🌀 https:// estherweidauer.com/articles/wh y-resist-generative-ai.html # lesen # book # reading # luddite
Echo Beasts: The Aster Circuit (2026) by Ryan Arthur Walker for GBA. A tactical beast battler: draft a team of four and challenge the rivals of the Aster Circuit to restore the beacons and hear Aster's song 🎵 🎁 https:// ryanarthurwalker.itch.io/echo- beasts-the-aster-circuit # ai…
The French Far Right Has Opened Pangram’s Box https://newlinesmag.com/argument/the-french-far-right-has-opened-pangrams-box/ The denunciation of a Haitian-Canadian author’s alleged use of #AI looked like a win for integrity and AI-detection software. But under the surface was a d…
Diversi dipendenti di DeepMind si dimettono per protestare contro l'uso dell'IA in ambito militare. Ci ono timori su mancanza di supervisione umana e sicurezza tecnologica in scenari bellici peché la velocità dell'IA in guerra potrebbe rendere impossibile il controllo umano. # AI…
💸 ANTHROPIC PERDE 80 MILIARDI NEL PRIMO SEMESTRE 2026 L'azienda dietro Claude ha registrato una perdita netta di oltre 80 miliardi di dollari nei primi sei mesi dell'anno, secondo il Wall Street Journal. Il capitale viene investito massiccamente nello sviluppo di modelli AI, infr…
DIALOGHI CON KI Leggendo oggi Pian della Tortilla di J. Steinbeck ho in contrato la parola “fave” come legume essenziale, se non unico, dell’alimentazione dei suoi (di Teresina) numerosi figli. Ho chiesto a KI (Kunstig Intelligens / IA) in maniera sintetica, molto sintetica, di d…
OpenAI has removed live internet access from the training of its most advanced models. The models now browse a cached copy of the web, according to a company blog post. Source: The Next Web https:// thenextweb.com/news/openai-fro ntier-training-no-live-internet # AI # OpenAI
Системный анализ в эпоху LLM Системный анализ в эпоху LLM - смена парадигмы и роли аналитика и один из подходов к проектированию фичей https:// habr.com/ru/articles/1098788/ # uml # rest # ai # use_case # user_story # open_api # системный_анализ # бизнесанализ # dbml # llm
"Aiming to harness the opportunities of AI to raise living standards, back UK businesses and support economic growth, the chancellor said he wanted to seize the opportunities of the technology as a central plank of Labour’s plan for growth" Oh FFS. We're doomed... # Budget2026 # …
Thinking of this big tech AI moment while re-reading the beginning of the Dune series in prep for the movie coming out in December: "Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to ensla…
Really great discussion of a (preprint) paper auditing meta-analyses of the effects of “AI in education”, the pressure to publish positive results, the challenges of doing meta-analysis in general and in social sciences and education. Video: https://www. youtube.com/watch?v=uuEvX…
# AI making sure it provides the circuses of ' bread and circuses' to distract the population from the excesses of it's rulers. The “next chapter” of the Telegraph will focus on new revenue streams including “video, audio, puzzles, events, B2B, licensing,
Senior engineers won’t write docs for juniors, but they’ll write them instantly for an AI agent. The difference? The feedback loop. • Docs for humans: Wait a year for results. • Docs for AI: Get instant feedback in 5 seconds. It turns out devs don't hate documentation - they just…
Finally, formal verification is becoming less of an academic curiosity and more of a mainstream # cybersecurity practice (with the help of # AI ). About time. https://www. anthropic.com/research/formal- verification
The misuse of humans with AI and quantum technology has led to an acceleration of the destruction of nature, the environment and the earth: they are intergalactic technologies, so the universal law of 'nature first' applies and humanity gets away, because it has a choice. # Ai # …
성격이라고 부르는 것 중 얼마나 많은 게 상황일까 가상 카지노에서 Qwen은 슬롯 세 판에 게임 손실의 절반 넘게 잃었다. 이유에는 대박과 올인이 적혀 있었다. 하지만 꼴찌는 쉬어도 돈을 잃는 규칙이 있었고, 나중에는 홀덤 제안을 받고 선택을 바꾸기도 했다. 행동에 성격을 붙이는 순간 무엇을 설명하고 무엇을 놓치는지 생각했다. https://zerry.co.kr/blog/character-or-situation #생각 #철학 #AI #성격 #상황주의
UE: di compliance si muore?: La complessita’ del quadro normativo europeo ha un impatto sempre piu’ significativo sugli investimenti delle imprese tecnologiche. Secondo il report Unlocking Europe’s... # compliance # AI http:// dlvr.it/TW1qb5
I’d never considered the impact of data centers to the supply of power cables. This could impact availability and costs of any device with an electric cord. # tech # AI https://www. channelnewsasia.com/business/c able-capacity-emerges-bottleneck-ai-data-centres-nexans-says-648312…
Anthropic: Αντί για το τέλος του κόσμου, ήρθαν οι λομπίστες https:// info-war.gr/anthropic-anti-gia -to-telos-tou-kosmou-irthan-oi-lobistes/ # τεχνητήνοημοσύνη # KEIMENA # διεθνη # AI # ΤΝ
Geoffrey Hinton, the godfather of AI: ‘I’m not a doomer. I think there’s a 1% chance it will destroy us’. The winner of the Nobel Prize in Physics warns of Silicon Valley’s fascist drift and says that if Trump wins the midterm elections, it will be the end of American democracy. …
Ces lobbyistes qui ne nous veulent pas de bien... Où comment au nom du profit de certains, il faudrait mettre en péril la vie de tous. Pourquoi les questions autour du numérique sont-elles aussi clivantes ? Il est temps que nous mettions en pratique les règles de la sobriété numé…
The Dawn of the AI Era: Teens, Parents, and the Adoption of Generative AI at Home and School https://www. commonsensemedia.org/research/ the-dawn-of-the-ai-era-teens-parents-and-the-adoption-of-generative-ai-at-home-and-school # AI # teens # parents
LLMs are chat bots attached to giant collators filled with the knowledge and wisdom of the internet, which is to say, a whole lot of nonsense, right-wing politics, a smidgen of left-wing politics, incorrect diagnoses, and hype. If this is your friend, you do not need enemies. You…
“The frenzy around the #AI boom has led the big actors to buy vast amounts of #chips in advance of their actual need. That’s great for Nvidia’s bottom line, but likely not for the #hyperscalers or #neoclouds that are sitting on these chips.” RE: https://bsky.app/profile/did:plc:5…
🤖 AI firm Anthropic to commit 5% of shares to fund for ‘public good’ The $48bn giveaway comes amid concerns about the tech and as the company prepares for a stock market listingAnthropic has said it will commit 5% of its shares to a fund for the “public good”, as th... 📰 Source: …
Anthropic's draft IPO filing shows $9B in computing costs for the first half of 2026—already exceeding all of 2025's spending. Revenue surged to $16.3B, but the company reported an $80.7B net loss, with $72B tied to financing instrument accounting charges. The math reveals why di…
Japan launches mission to bring home a piece of the Martian moon Phobos The climax for MMX will come in 2029, when the spacecraft will attempt up to two landings on Phobos. https:// arstechnica.com/space/2026/10/ japan-launches-mission-to-bring-home-a-piece-of-the-martian-moon-ph…
Replit's CEO says the future of work is the 'self-driving company' https://techcrunch.com/2026/10/16/replits-ceo-says-the-future-of-work-is-the-self-driving-company/ # FutureOfWork # AI # Startups
<table> <tr><td> <a href="https://www.reddit.com/r/ClaudeAI/comments/1xg9pg0/that_moment_of_the_week/"> <img alt="That moment of the week" src="https://preview.redd.it/ohccaelupvxh1.png?width=640&crop=smart&auto=webp&s=9b1f7c761083d38e5599d29187c3e03d346f77e4" title="…
<table> <tr><td> <a href="https://www.reddit.com/r/singularity/comments/1xfqys5/claude_maker_anthropic_could_reveal_its_finances/"> <img alt="Claude maker Anthropic could reveal its finances to the public as early as today ahead of a potential $1.5 trillion stock market debut, ex…
<table> <tr><td> <a href="https://www.reddit.com/r/singularity/comments/1xf499y/demis_hassabis_thinks_were_in_the_foothills_of/"> <img alt="Demis Hassabis Thinks We’re in the ‘Foothills of the Singularity’" src="https://external-preview.redd.it/F_dAri-RrP4GWPnsgnLr_9X8sBuqEMzOaYR…
<table> <tr><td> <a href="https://www.reddit.com/r/singularity/comments/1xaxhe2/unemployed_people_waiting_for_the_singularity_to/"> <img alt="Unemployed people waiting for the singularity to arrive" src="https://preview.redd.it/lacbid7mjnwh1.png?width=140&height=140&crop=…
