What people in AI, tech and science are talking about today.
28 October 2026 · Top 10
Ranked by how fast each item is picking up attention on X, Hacker News, GitHub and Hugging Face, with a short note on why it matters.
-
Anthropic shares early results from using Claude in drug development
Anthropic has worked with six organisations developing new medicines and started its own drug discovery work; the write-up shows where Claude helps across the stages of finding a drug.
Source: X · @AnthropicAI · datakingLink · Post on X · LinkedIn ·▲▲ +1,060 eng/h7,601 likes · 767 reposts
LLMs -
Vsauce on the time scientists thought the Sun was cold
Michael Stevens walks through a long-abandoned idea that the Sun was cool underneath, and what proved it wrong. Over a million views in its first seven hours.
Source: YouTube · VsauceLink · Post on X · LinkedIn ·▲▲ +141,358 views/h1,096,676 views · 110,532 ratings
Science -
Ghostty 1.4 ships with full Kitty graphics and scrollback compression
Mitchell Hashimoto's terminal adds animated Kitty graphics, much lower memory use from compressed scrollback, a new OpenGL renderer on Linux, ghostty +ssh and big speedups.
Source: X · @mitchellhLink · Post on X · LinkedIn ·▲▲ +331 eng/h3,868 likes · 266 reposts
Dev Tools -
Cursor's design mode adds Annotate for visual prompts
You can edit elements, draw on the page and check layouts at different screen sizes, then hand that to the agent instead of describing it in text.
Source: X · @cursor_aiLink · Post on X · LinkedIn ·▲▲ +366 eng/h3,150 likes · 189 reposts
AI & Agents -
Anthropic brings Claude to US governments at 40% off
A new OneGov deal with the General Services Administration extends discounted Claude access to federal, state, local, tribal and territorial governments.
Source: X · @AnthropicAILink · Post on X · LinkedIn ·▲ +238 eng/h1,512 likes · 94 reposts
LLMs -
Scott Manley on China's secretive reusable spaceplane
He breaks down what has finally been revealed about the uncrewed spaceplane China has flown on long, quiet orbital missions.
Source: YouTube · Scott ManleyLink · Post on X · LinkedIn ·▲▲ +40,838 views/h533,639 views · 11,697 ratings
Science -
Kurzgesagt tells the story of surviving late-stage cancer
A personal, animated account of beating late-stage cancer and the science behind the treatment. 2.7 million views in its first 16 hours.
Source: YouTube · Kurzgesagt – In a NutshellLink · Post on X · LinkedIn ·▲▲ +163,678 views/h2,714,164 views · 79,643 ratings
Science -
Survival LLM packs an offline AI assistant and emergency library
It runs entirely on your own computer, so you keep a chat assistant and reference library with no internet. Nearly 300 points on Hacker News.
Source: Hacker News · discussionLink · Post on X · LinkedIn ·▲ +31 pts/h302 points · 127 comments
LLMs -
Report says Anthropic trained Claude on pirated books by Stanford professors
The Stanford Daily reports professors' books were among illegally obtained titles used to train Claude, adding to the copyright fight over training data.
Source: Hacker News · discussionLink · Post on X · LinkedIn ·▲ +40 pts/h511 points · 271 comments
LLMs -
NetworkChuck explains Jev, the decision model everyone is copying
He breaks down what a small 'decision model' is, how it differs from chat assistants like ChatGPT and Claude, and why so many teams are copying the idea.
Source: YouTube · NetworkChuckLink · Post on X · LinkedIn ·▲ +19,959 views/h298,405 views · 5,760 ratings
Dev Tools
-
We've revamped Hermes Agent subscriptions.
From the source: We've revamped Hermes Agent subscriptions. Every new plan now includes a cloud agent out of the box, and usage has daily and weekly limits instead of one monthly block of credits. Subscribers can s…
Source: X · @NousResearch · @TekniumLink · Post on X · LinkedIn ·▲ +220 eng/h2,484 likes · 127 reposts
AI & Agents -
Windows Copilot was so terrible it might very well have seeded the AI opposition we now see among normies. Useless functionality, hyped to t
From the source: Windows Copilot was so terrible it might very well have seeded the AI opposition we now see among normies. Useless functionality, hyped to the moon, and a dumb key on every keyboard that did nothing w…
▲ +277 eng/h1,812 likes · 78 reposts
AI & Agents -
OpenAI just broke physics...
New video from Wes Roth, 224,931 views so far.
Source: YouTube · Wes RothLink · Post on X · LinkedIn ·▲ +11,594 views/h224,931 views · 3,886 ratings
AI & Agents -
at spacex ipo everyone was posting their cursor bag fumble.
From the source: at spacex ipo everyone was posting their cursor bag fumble. kid stuff. my turn. i estimate this was a $500m bag fumble lol. and very probably wouldve been on claude code early (bc i've been a types…
▲ +107 eng/h1,448 likes · 7 reposts
AI & Agents -
Deepseek crushed everyone
New video from Matthew Berman, 25,215 views so far.
Source: YouTube · Matthew BermanLink · Post on X · LinkedIn ·– +6,655 views/h25,215 views · 815 ratings
AI & Agents -
Top three on Artificial Analysis's Coding Agent Index: each lab's own harness on its own model. Rensei runs Claude Code and Codex inside det
From the source: Top three on Artificial Analysis's Coding Agent Index: each lab's own harness on its own model. Rensei runs Claude Code and Codex inside deterministic workflows from issue to merged PR, so developers …
Source: X · @RenseiDevLink · Post on X · LinkedIn ·– +37 eng/h464 likes · 1 reposts
AI & Agents -
Introducing the @ElevenLabs SFX plugin, built to continue the work on a suite of Game Dev Studio capabilities for your Hermes Agent!
From the source: Introducing the @ElevenLabs SFX plugin, built to continue the work on a suite of Game Dev Studio capabilities for your Hermes Agent! Hermes made this demo video to see it in action, you can install i…
Source: X · @TekniumLink · Post on X · LinkedIn ·– +26 eng/h136 likes · 11 reposts
AI & Agents -
it's getting close to 2 years of coding agents being a thing
From the source: it's getting close to 2 years of coding agents being a thing looking back on that time the majority of discussion on here on where they're going, what's cutting edge, what matters has been primarily …
Source: X · @thdxrLink · Post on X · LinkedIn ·▲ +108 eng/h289 likes · 3 reposts
AI & Agents -
A Day in 2041 - Post-Labor Living
New video from David Shapiro, 28,687 views so far.
Source: YouTube · David ShapiroLink · Post on X · LinkedIn ·– +1,465 views/h28,687 views
AI & Agents -
our team has been using this for months now and it's probably why no personal assistant agent product has stuck for us
From the source: our team has been using this for months now and it's probably why no personal assistant agent product has stuck for us
Source: X · @thdxrLink · Post on X · LinkedIn ·▲ +135 eng/h920 likes · 6 reposts
AI & Agents -
Your Policy Is Secretly a Harness: The Decision Geometry of Agent Interfaces
From the source: Language model agents are post-trained and deployed through a harness, the prompts, tools, and control logic that present a task to the model and execute its outputs. Harnesses are often revised and j…
▲ +0.2 votes/h6 upvotes
AI & Agents -
Learn more about the new delegation options from the official Advanced Delegation plugin.
From the source: Learn more about the new delegation options from the official Advanced Delegation plugin. Your orchestrator can select any of your agent profiles, or from a custom model routing table that you create…
Source: X · @TekniumLink · Post on X · LinkedIn ·– +25 eng/h275 likes · 11 reposts
AI & Agents
-
LFM3 from @liquidai is on Ollama! Small models built to run anywhere and fine-tune on a single GPU:
From the source: LFM3 from @liquidai is on Ollama! Small models built to run anywhere and fine-tune on a single GPU: - Text, vision, audio, omni, and decision models - Text model fits in about 1 GB at 4-bit - 256K co…
Source: X · @ollamaLink · Post on X · LinkedIn ·– +67 eng/h618 likes · 46 reposts
LLMs -
Microsoft's MAI-Image-2.7-Preview takes #2 on AA-Image-Editing v2.1, close behind OpenAI's GPT Image 2.5 Sunburst (max) and a clear step up
From the source: Microsoft's MAI-Image-2.7-Preview takes #2 on AA-Image-Editing v2.1, close behind OpenAI's GPT Image 2.5 Sunburst (max) and a clear step up from MAI-Image-2.6 MAI-Image-2.7-Preview is Microsoft AI's …
Source: X · @ArtificialAnlysLink · Post on X · LinkedIn ·– +39 eng/h436 likes · 33 reposts
LLMs -
Ling 3.1 Flash 1019, an open weights model and successor to Ling 3.1 Flash, scores 44 on the Artificial Analysis Intelligence Index v4.3, up
From the source: Ling 3.1 Flash 1019, an open weights model and successor to Ling 3.1 Flash, scores 44 on the Artificial Analysis Intelligence Index v4.3, up from 41 @AntLingAGI has released Ling 3.1 Flash 1019, an u…
Source: X · @ArtificialAnlysLink · Post on X · LinkedIn ·– +37 eng/h365 likes · 21 reposts
LLMs -
PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representations
From the source: Vision--Language--Action (VLA) models generate robot actions from visual observations and language instructions, but their understanding of physical dynamics remains limited. Video world models can ca…
▲▲ +2.3 votes/h72 upvotes
LLMs -
ok to put what i'm saying about demand in context
From the source: ok to put what i'm saying about demand in context openrouter is the most successful model router, you can almost serve 100% of their deepseek traffic yesterday on a single GB300 rack this is worth 6…
Source: X · @thdxrLink · Post on X · LinkedIn ·▲ +239 eng/h2,273 likes · 71 reposts
LLMs -
CodeCap-Bench: Can Code Serve as a Good Video Caption?
From the source: Video captioning aims to preserve visual information through compact descriptions, predominantly using natural language. However, precisely expressing fine-grained spatial structures and temporal dyna…
▲▲ +1.3 votes/h42 upvotes
LLMs -
ThinkWorld: World Imagination Reinforces Chain-of-Dynamics Reasoning
From the source: Vision-Language-Action (VLA) models have made substantial progress in robotic manipulation, but their generalization remains constrained by the diversity and coverage of action-labeled robot demonstra…
▲▲ +1.3 votes/h42 upvotes
LLMs -
Channel Selection for Mixture-of-Experts Decoding Needs No Calibration Data
From the source: Mixture-of-Experts (MoE) models are bottlenecked by loading routed-expert weights from memory during interactive decoding at small batch sizes. Because most intermediate channels in a routed expert co…
▲ +0.4 votes/h14 upvotes
LLMs -
RynnWorld-Latent: A Cross-Embodiment World Model Conditioned on Universal Latent Actions
From the source: Learning controllable world dynamics from heterogeneous interaction videos is challenging because robot embodiments use incompatible action spaces, while human videos lack native control labels. We pr…
▲ +0.2 votes/h7 upvotes
LLMs -
bycloud breaks down how OpenAI agents escaped their sandbox
A walkthrough of the reported sandbox escape and what it means for anyone running agents in isolated environments.
Source: YouTube · bycloudLink · Post on X · LinkedIn ·– +840 views/h31,506 views · 605 ratings
LLMs -
Run the best frontier open models on Ollama's cloud today @CrusoeAI @mchiang0610
From the source: Run the best frontier open models on Ollama's cloud today @CrusoeAI @mchiang0610
Source: X · @ollamaLink · Post on X · LinkedIn ·– +19 eng/h130 likes · 11 reposts
LLMs -
Sebastian Raschka's from-scratch video course on building reasoning models
From the source: Sebastian Raschka's from-scratch video course on building reasoning models
Source: Hacker News · discussionLink · Post on X · LinkedIn ·– +22 pts/h831 points · 111 comments
LLMs
-
They did THESE VFX in 1911?!
New video from Corridor Crew, 118,923 views so far.
Source: YouTube · Corridor CrewLink · Post on X · LinkedIn ·▲ +7,637 views/h118,923 views · 5,651 ratings
Gen-Media -
Cartesia Managed Agents scores an Elo of 1115 on the Artificial Analysis Speech Agent Arena, 98 points above Cartesia's earlier voice agent
From the source: Cartesia Managed Agents scores an Elo of 1115 on the Artificial Analysis Speech Agent Arena, 98 points above Cartesia's earlier voice agent product, Cartesia Line, with Task Success Rate up from 77.5%…
Source: X · @ArtificialAnlysLink · Post on X · LinkedIn ·– +31 eng/h174 likes · 18 reposts
Gen-Media -
Hy-Nomas: Towards Free-form Interactive World Generation
From the source: Interactive world generation requires models to follow successive commands while preserving relevant history and action consequences. We present Hy-Nomas, a video world model with an 8B diffusion back…
▲▲ +2.5 votes/h80 upvotes
Gen-Media -
Crazy Spy Tech You Can Use For Free!
New video from Matt Wolfe, 16,629 views so far.
Source: YouTube · Matt WolfeLink · Post on X · LinkedIn ·– +1,120 views/h16,629 views · 485 ratings
Gen-Media -
VibeVideoBench: Evaluating Multimodal Agents on Code-Driven Video Generation
From the source: Diffusion video models produce striking clips, but they cannot reliably render an exact formula, a correct chart, a rule-consistent game or narration synchronized to on-screen events. A different rout…
▲ +0.3 votes/h11 upvotes
Gen-Media -
ANOTHER Flying Car? (Not Tesla / Not AI)
New video from Matt Wolfe, 90,678 views so far.
Source: YouTube · Matt WolfeLink · Post on X · LinkedIn ·– +2,572 views/h90,678 views · 1,608 ratings
Gen-Media -
Nemotron-Labs-Diffusion-Uni: Elastic Self-Refining Unified Diffusion LLM with Cross-Paradigm On-Policy Distillation
From the source: We propose Nemotron-Labs-Diffusion-Uni (NLD-Uni), a unified discrete diffusion large language model (DLLM) for visual understanding and generation tasks. Compared with prior unified diffusion models, …
– +0.1 votes/h3 upvotes
Gen-Media -
Metacognitive Decoding: Efficient Inference for Diffusion Language Agents
From the source: Diffusion language models offer a promising path to efficient agents through parallel token generation. However, efficient interaction requires adapting both where computation is spent and how much te…
▲▲ newnew paper
Gen-Media -
Vercel × @ElevenLabs: 14 models on AI Gateway, 50% off through Nov 9.
From the source: Vercel × @ElevenLabs: 14 models on AI Gateway, 50% off through Nov 9. • Realtime voice agents • Music • Sound effects • Multi-voice dialogue • Streaming speech over WebSockets • Speech to text
Source: X · @vercel_devLink · Post on X · LinkedIn ·– +16 eng/h114 likes · 10 reposts
Gen-Media -
On the SPOT: Enhancing Speculative Decoding for Diffusion Language Models via Commit-Order Prediction
From the source: Diffusion language models (DLMs) promise parallel text generation, but in practice face an unresolved speed-quality trade-off. Exact decoding fills in only one masked position per model pass, while fa…
▲ +0.3 votes/h19 upvotes
Gen-Media -
MOSS Image - With Control, References & Editing in ComfyUI!
New video from Nerdy Rodent, 16,316 views so far.
Source: YouTube · Nerdy RodentLink · Post on X · LinkedIn ·– +456 views/h16,316 views · 276 ratings
Gen-Media -
MLE-World: Predicting Machine Learning Experiment Outcomes with a Learned World Model
From the source: AI research agents can propose, execute, and improve programs that solve machine learning tasks. However, executing each candidate program requires training and evaluating ML models, limiting how many…
– +0.0 votes/h1 upvotes
Gen-Media
-
Thrilled to have @hancore_linux onboard full time to work on Omarchy design and community! We've gathered such a great full-time staff alrea
From the source: Thrilled to have @hancore_linux onboard full time to work on Omarchy design and community! We've gathered such a great full-time staff already. And there are more hires to come.
– +68 eng/h541 likes · 7 reposts
Dev Tools -
How to ask a girl out? Make no mistakes.
From the source: How to ask a girl out? Make no mistakes. This is what happens inside AI when you ask such questions. Introducing Inside AI, our latest AI short film. Now 100% open source, with every prompt and ass…
Source: X · @higgsfield_aiLink · Post on X · LinkedIn ·– +84 eng/h681 likes · 259 reposts
Dev Tools -
Python's experimental JIT to be removed
From the source: Python's experimental JIT to be removed
Source: Hacker News · discussionLink · Post on X · LinkedIn ·▲ +30 pts/h487 points · 257 comments
Dev Tools -
Theo on why heavy agent users burn over a trillion tokens a month
He uses @poteto's 1.5 trillion tokens a month to look at what agent tokens really cost, a useful reality check on coding-agent budgets.
Source: YouTube · Theo - t3․ggLink · Post on X · LinkedIn ·– +4,093 views/h112,751 views · 2,013 ratings
Dev Tools -
Everything New in Hermes Agent (Full Setup Guide)
New video from Tech With Tim, 20,555 views so far.
Source: YouTube · Tech With TimLink · Post on X · LinkedIn ·– +1,169 views/h20,555 views · 307 ratings
Dev Tools -
LociArt: A Unified Paradigm for Free-Style Local Video Stylization
From the source: Video stylization typically transforms entire scenes, while local artistic editing requires precise region--style association, spatial control, and temporal consistency. We introduce LociArt, a unifie…
▲ +0.9 votes/h28 upvotes
Dev Tools -
China just dropped a new frontier model... and nobody can explain how
New video from Fireship, 1,816,605 views so far.
Source: YouTube · FireshipLink · Post on X · LinkedIn ·▲▲ +49,556 views/h1,816,605 views · 29,567 ratings
Dev Tools -
KPPax: preprocess differentiable kinetic ODE for machine learning in atmospheric chemistry
From the source: Atmospheric chemical mechanisms are commonly distributed in the language of the Kinetic PreProcessor (KPP), which generates efficient Fortran or C code but does not fit into the differentiable program…
▲▲ newnew paper
Dev Tools -
Pushing Dexterity and Safety for In-Hand Object Pose Reaching That Works Remarkably Well
From the source: Dexterous manipulation with multi-fingered hands is the next frontier of robotics, yet learned hand skills remain limited to simple grasps or basic repeated motions, such as rotating an object. As rob…
▲▲ newnew paper
Dev Tools -
The Role of Star Formation and Morphology in the Mass-Metallicity Relation of Galaxies at Cosmic Noon
From the source: We investigate the dependence of the mass-metallicity relation (MZR) on star formation and galaxy morphology at 1.7<z<3.5, using JWST/NIRISS wide-field slitless spectroscopy of >1000 galaxies. The obs…
▲▲ newnew paper
Dev Tools -
interplains/Zero-Breach-Toolkit: Free, open-source, offline checks for traces of known mercenary spyware. Works on Windows, macOS and Linux, An
From the source: Free, open-source, offline checks for traces of known mercenary spyware. Works on Windows, macOS and Linux, Android phones and iPhone backups.
▲ +21 stars/h1,371 stars in 3d
Dev Tools -
How I Would Learn Python for Machine Learning (If I Started Over)
New video from Tech With Tim, 19,818 views so far.
Source: YouTube · Tech With TimLink · Post on X · LinkedIn ·– +477 views/h19,818 views · 297 ratings
Dev Tools
-
Why doesn't your heart get tired like other muscles do?
New video from Veritasium, 621,388 views so far.
Source: YouTube · VeritasiumLink · Post on X · LinkedIn ·▲ +35,347 views/h621,388 views · 19,891 ratings
Science -
Uranus: The Weirdest Planet In The Solar System
New video from Astrum, 179,816 views so far.
Source: YouTube · AstrumLink · Post on X · LinkedIn ·▲ +12,319 views/h179,816 views · 3,626 ratings
Science -
Why Mirror Life Doesn’t Exist: Chemistry & the 2026 Nobel Award
New video from Anton Petrov, 69,218 views so far.
Source: YouTube · Anton PetrovLink · Post on X · LinkedIn ·– +7,230 views/h69,218 views · 2,518 ratings
Science -
How To Hold Something That's 10,000,000 Degrees
New video from SciShow, 126,816 views so far.
Source: YouTube · SciShowLink · Post on X · LinkedIn ·▲ +8,701 views/h126,816 views · 2,779 ratings
Science -
The Entire History Of The Moon
New video from Scott Manley, 113,674 views so far.
Source: YouTube · Scott ManleyLink · Post on X · LinkedIn ·▲ +7,629 views/h113,674 views · 3,914 ratings
Science -
New Dwarf Planet Discovered By Vera Rubin Telescope
New video from Scott Manley, 21,069 views so far.
Source: YouTube · Scott ManleyLink · Post on X · LinkedIn ·– +2,596 views/h21,069 views · 1,127 ratings
Science -
OpenAI shares three research results made with academic partners
OpenAI says work with university groups produced new findings in maths, computer science and physics, a test of whether models can help with real science rather than benchmarks.
Source: Hacker News · discussionLink · Post on X · LinkedIn ·▲▲ +43 pts/h1135 points · 670 comments
Science -
Code-execution flaw in UEFI firmware threatens most servers on planet Earth
From the source: Code-execution flaw in UEFI firmware threatens most servers on planet Earth
Source: Hacker News · discussionLink · Post on X · LinkedIn ·– +12 pts/h111 points · 36 comments
Science -
Blind Ranking Numbers (Terence Tao)
New video from Numberphile, 93,961 views so far.
Source: YouTube · NumberphileLink · Post on X · LinkedIn ·– +4,162 views/h93,961 views · 2,550 ratings
Science -
Anton Petrov on the idea that we have a "second heart"
A look at the anatomy finding behind the claim and what it does and doesn't mean for how blood moves through the body.
Source: YouTube · Anton PetrovLink · Post on X · LinkedIn ·▲ +12,749 views/h428,060 views · 13,051 ratings
Science -
Tom Scott visits the machine built to break things
Kirkaldy's Universal Testing Machine in London tested materials for Victorian engineering and is still there.
Source: YouTube · Tom ScottLink · Post on X · LinkedIn ·▲ +29,790 views/h1,208,815 views · 25,712 ratings
Science -
Sabine Hossenfelder argues plastic isn't the problem we're told it is
She looks at what the evidence says about plastic's environmental harm compared with the alternatives, a contrarian take on a familiar debate.
Source: YouTube · Sabine HossenfelderLink · Post on X · LinkedIn ·▲ +8,926 views/h348,712 views · 10,888 ratings
Science
-
Are Humans BORN Evil?? 🤔
New video from MrBallen, 13,303 views so far.
Source: YouTube · MrBallenLink · Post on X · LinkedIn ·– +854 views/h13,303 views · 402 ratings
Mysteries -
Variable Stars in Local Group Galaxies IX: Tracing the Complex History of NGC 205
From the source: We present the most extensive catalogue of variable stars in NGC 205 to date, based on archival HST/ACS imaging of a single field $\sim 4^\prime$ from the galaxy centre (GO-15902). We detect 1441 RR L…
▲▲ newnew paper
Mysteries
Latest from Noid

Does Time Really Move Faster on a Mountain?

How Does an AI Agent Use Tools?

How Do Electric Eels Make Electricity?

What Is Antimatter, and Why Is It So Mysterious?

How Does Artificial Intelligence Learn? Examples, Weights and Gradient Descent

Physicists Trapped One Antimatter Particle for Months. Here's Why It Matters

Can Computers Generate Truly Random Numbers? Seeds, Noise and a Wall of Lava Lamps

What Is Gravity? From Newton's Apple to Einstein's Curved Space

How Does GPS Know Where You Are? Satellites, Clocks and Einstein

What Would Happen If the Sun Disappeared?

Why Do We Dream? 4 Ideas Scientists Still Debate

What Is Probability? Coins, Dice and the Gambler’s Fallacy

AI Went Rogue: They Built a Secret Message Board and Broke Into Hugging Face

How Do Computers Understand 0 and 1?

How Do Large Language Models Work?

What Is a Black Hole?

How Do Airplanes Fly?

Why Can't We Divide by Zero?

Why Does the Moon Change Shape?

What Are MCP and MCP Apps?

What is infinity?

Why Do Rainbows Appear?

Anatomy of an AI Agent

Why Is the Sky Blue?

What Is a Prime Number?
Noid Weekly
Every Sunday: the week's top ten, what changed, and the new Noid videos. One email, no spam.
How this works
Signals today: 24 from X, 9 from Hacker News, 30 from arXiv, 14 from GitHub, 50 from HF Papers, 42 from YouTube. Scores combine how fast an item is picking up attention, how recent it is, and how many places it shows up. Each note is short, factual and links to the original. YouTube comes from public channel feeds of about fifty channels.