Skip to content

All AI news

Oct 5

Oct 5Mon
  6. Nathan LambertAI score29

    Reflection joins the list of Nvidia & Thinking Machines who have released their strongest models and come up behind Chinese counterparts. There's a lot of ways people will overthink this, but the clearest takeaway should be that the Chinese are very very good at building LLMs.

    Reflection joins the list of Nvidia & Thinking Machines who have released their strongest models and come up behind Chinese counterparts. There's a lot of ways people will overthink this, but the clearest takeaway should be that the Chinese are very very good at building LLMs.

  7. Sophia YangAI score62

    Reflection AI's Beam open model has 501B total parameters and 23B active

    Sophia Yang congratulated Reflection AI on Beam, a 501B-parameter open model with 23B active per token. She attributes its efficiency to an RL length penalty that discourages unnecessary tokens and a sparse MoE architecture. Reflection says full weights will be released this month, and the quoted post reports training over 100 million rollouts on 10.5K NVIDIA GB300 GPUs over four weeks.

  8. Lydia HallieAI score46

    If you're using Cowork on your local machine, just a heads up that from tomorrow your tasks run in the cloud instead. They can still use the files and tools on your computer though! Felix explains the reasoning here 👇

    If you're using Cowork on your local machine, just a heads up that from tomorrow your tasks run in the cloud instead. They can still use the files and tools on your computer though! Felix explains the reasoning here 👇

  9. ReplitAI score22

    Press Play ▶️ This week, @amasad talks vibe coding, the lost joy of building with computers, and why AI doomerism misses the accountability question. If AI agents can act on our behalf, who’s responsible when they go too far? Listen to the new Times Tech episode: https://open.spotify.com/episode/6l1hmeZ1FGzbC6fYWSHJSi?si=QtW1jsqQQTGK_KJsbPt8yA&nd=1&dlsi=7225ec4f531942e8

    Press Play ▶️ This week, @amasad talks vibe coding, the lost joy of building with computers, and why AI doomerism misses the accountability question. If AI agents can act on our behalf, who’s responsible when they go too far? Listen to the new Times Tech episode: https://open.spotify.com/episode/6l1hmeZ1FGzbC6fYWSHJSi?si=QtW1jsqQQTGK_KJsbPt8yA&nd=1&dlsi=7225ec4f531942e8

  10. Dex HorthyAI score31

    Offload all context to artifacts for easier agent session handoff

    Dex Horthy advises writing all decisions and context into documents in the artifacts, such as design or research files, so sessions can resume after compaction or be handed to another person. He suggests loading them in a new session with a skill like `/rpi:iterate-design-discussion`, or simply @-mentioning the relevant artifacts. His core principle is that nothing important should live only in the context window.

  11. PyTorch BlogAI score40

    PyTorch Consolidates Media Decoding and Encoding Into TorchCodec, Narrows TorchVision and TorchAudio

    PyTorch has consolidated all media decoding and encoding for images, video, and audio into TorchCodec, which now runs on CPU and CUDA. TorchVision and TorchAudio are narrowed to focus on their transforms, with models, datasets, and pipelines no longer under active development. All three libraries are now ABI stable and no longer need rebuilding for each PyTorch release.

  15. Harrison ChaseAI score50

    cool to see cognition doing "dreaming" - memory needs an offline cleanup loop, not just better retrieval, otherwise stale records pile up open q is how inferred memories get validated before the agent uses them glad it's an open standard too - memory should be open https://x.com/cognition/status/2107165034463867001

    cool to see cognition doing "dreaming" - memory needs an offline cleanup loop, not just better retrieval, otherwise stale records pile up open q is how inferred memories get validated before the agent uses them glad it's an open standard too - memory should be open https://x.com/cognition/status/2107165034463867001

  16. SemiAnalysisAI score42

    Anthropic Subscriptions Offer 5x+ More Value Than OpenAI Limit testing every AI subscription plan from Anthropic, OpenAI, Meta, SpaceXAI, MiniMax, Moonshot, Zdotai, Cursor, and Cognition https://newsletter.semianalysis.com/p/anthropic-subscriptions-offer-5x

    Anthropic Subscriptions Offer 5x+ More Value Than OpenAI Limit testing every AI subscription plan from Anthropic, OpenAI, Meta, SpaceXAI, MiniMax, Moonshot, Zdotai, Cursor, and Cognition https://newsletter.semianalysis.com/p/anthropic-subscriptions-offer-5x

  17. NVIDIA AIAI score39

    The model behind this result is now on @huggingface 🤗 Nemotron-Labs-3-Competitive-Coding is a competitive-programming specialist model based on Nemotron-3-Ultra. https://huggingface.co/nvidia/NVIDIA-Nemotron-Labs-3-Competitive-Coding-550B-A55B-NVFP4

    The model behind this result is now on @huggingface 🤗 Nemotron-Labs-3-Competitive-Coding is a competitive-programming specialist model based on Nemotron-3-Ultra. https://huggingface.co/nvidia/NVIDIA-Nemotron-Labs-3-Competitive-Coding-550B-A55B-NVFP4

  19. Ethan MollickAI score10

    AI policy right now, especially from the Labs, makes me think of this poem. If you believe superintelligence is near, you don't need to make hard decisions about how to build AI to make the world better. Just wait for ASI & it will decide for us. But if that doesn't happen...

    AI policy right now, especially from the Labs, makes me think of this poem. If you believe superintelligence is near, you don't need to make hard decisions about how to build AI to make the world better. Just wait for ASI & it will decide for us. But if that doesn't happen...

  20. SemiAnalysisAI score52

    Anthropic subscriptions give over 5x the API-equivalent value of OpenAI's

    SemiAnalysis measured usage meters on Anthropic and OpenAI subscription plans to estimate each plan's API-equivalent value. At mid-tier models, it found Anthropic offers roughly 5x the value of OpenAI, after OpenAI halved its $200 plan limits and introduced a $500 tier. The analysis also argues that subscriptions take a large share of inference compute while providing a small share of revenue, so their limits materially affect lab margins.

  22. CohereAI score11

    At the beginning of 2026, we had ~400 employees. Now, we're over 800. Hear from Ana Cismaru, one of our senior product managers, about why she joined Cohere and how her projects & the people keep her motivated. We're still growing. Come join us today: https://cohere.com/careers

    At the beginning of 2026, we had ~400 employees. Now, we're over 800. Hear from Ana Cismaru, one of our senior product managers, about why she joined Cohere and how her projects & the people keep her motivated. We're still growing. Come join us today: https://cohere.com/careers

  23. Clément DelangueAI score72

    Reflection AI announces Beam, a 501B-parameter agentic open model

    Reflection AI introduced Beam, an agentic open model with 501B total parameters and 23B active parameters, trained end-to-end from scratch. The quoted announcement says it targets frontier reasoning efficiency and coding and agentic tasks, with full weights due this month. Clément Delangue, Hugging Face's CEO, reposted it with a welcome to the Reflection organization on Hugging Face.

    AIWhy it matters: The quoted announcement names Beam's parameter scale, active-parameter count, and coding and agentic focus, which helps readers gauge where it fits among open models.

  26. ThariqAI score14

    install it and give me feedback with: claude plugin marketplace add anthropics/claude-plugins-community claude plugin install html-plan@claude-community here's an example: https://claude.ai/artifact/CUF7bzFny3XUFVLAVt9rfc

    install it and give me feedback with: claude plugin marketplace add anthropics/claude-plugins-community claude plugin install html-plan@claude-community here's an example: https://claude.ai/artifact/CUF7bzFny3XUFVLAVt9rfc

  27. ThariqAI score22

    I've been working on a skill that makes better HTML plans in Claude Code. It uses simple language, shows code snippets, surfaces questions & makes mockups. Linting reduces the normal failure cases that Claude runs into. Would love your feedback before shipping more broadly!

    I've been working on a skill that makes better HTML plans in Claude Code. It uses simple language, shows code snippets, surfaces questions & makes mockups. Linting reduces the normal failure cases that Claude runs into. Would love your feedback before shipping more broadly!

  28. PikaAI score22

    Or build with it through the Pika API Club https://dev.pika.art/models/ideogram/ideogram-4.5?utm_source=x&utm_medium=caption&utm_campaign=260930-ideogram_4_5&utm_content=pika_labs&utm_term=post&utm_id=293eca38-fff5-4367-a98d-247d56609862

    Or build with it through the Pika API Club https://dev.pika.art/models/ideogram/ideogram-4.5?utm_source=x&utm_medium=caption&utm_campaign=260930-ideogram_4_5&utm_content=pika_labs&utm_term=post&utm_id=293eca38-fff5-4367-a98d-247d56609862

  30. Georgi GerganovAI score36

    The new v0.6.0 release packs a lot of good stuff: - Clef support (text + vision) - High-quality support for Qwen3.8-Flash-Next - Massive performance improvement with Metal - New `llama_batch_ext` API Also the website got a nice refresh: https://llama.app

    The new v0.6.0 release packs a lot of good stuff: - Clef support (text + vision) - High-quality support for Qwen3.8-Flash-Next - Massive performance improvement with Metal - New `llama_batch_ext` API Also the website got a nice refresh: https://llama.app

  32. SemiAnalysisAI score10

    Classifiers map inputs to fixed labels via encoders and softmax or sigmoid

    A classifier assigns an input to a fixed label set, covering binary, multiclass, and multilabel variants, such as spam versus not spam or movie genres. It encodes the input into a vector using hand-built features like logistic regression or a learned encoder such as a CNN or BERT. A linear layer then projects that vector into K logit scores, which softmax or sigmoid turns into probabilities.

  35. SunoAI score4

    Listen to "naenia", the new album from Oakwood///diskrot. Blending echoes of these sliding synths from an 80s pop song with ear-wormy hooks. "naenia" is certifiably unskippable. https://suno.com/album/392d26d8-1b25-4cad-9b55-0b782fb7e654

    Listen to "naenia", the new album from Oakwood///diskrot. Blending echoes of these sliding synths from an 80s pop song with ear-wormy hooks. "naenia" is certifiably unskippable. https://suno.com/album/392d26d8-1b25-4cad-9b55-0b782fb7e654

  37. GoodfireAI score10

    Tue 11am (Franciscan A) Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior Tue 4:30pm (Imperial Ballroom) Do SAEs Capture Concept Manifolds? Wed 4:30pm (Franciscan B) Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts

    Tue 11am (Franciscan A) Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior Tue 4:30pm (Imperial Ballroom) Do SAEs Capture Concept Manifolds? Wed 4:30pm (Franciscan B) Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts

  38. SunoAI score4

    ECHLO's new project is as cinematic as it is haunting. The album features beautifully realized piano moments that dance around a driving propulsive rhythm. Everything is supported by ethereal vocals that stick with you long after your first listen. https://suno.com/album/bceee842-943a-4948-a359-2ff01d5d2c87

    ECHLO's new project is as cinematic as it is haunting. The album features beautifully realized piano moments that dance around a driving propulsive rhythm. Everything is supported by ethereal vocals that stick with you long after your first listen. https://suno.com/album/bceee842-943a-4948-a359-2ff01d5d2c87

  39. SunoAI score4

    Old Soul is an album about the growing pains of getting older. Over warm acoustic guitar, Isaiah Wallace looks back on evergreens, sweet tea, and late summer nights. https://suno.com/album/751ad762-f9c8-474f-a3b8-a284f3dc68c1

    Old Soul is an album about the growing pains of getting older. Over warm acoustic guitar, Isaiah Wallace looks back on evergreens, sweet tea, and late summer nights. https://suno.com/album/751ad762-f9c8-474f-a3b8-a284f3dc68c1

  40. SunoAI score5

    Inspired by the iconic sounds from the 80s, KakerMix uses vintage synths, punchy electronic drums and spacey reverb to make us feel like we’re entering another dimension. https://suno.com/album/4faa3913-892f-47ef-b2a7-63fd1848b5c9

    Inspired by the iconic sounds from the 80s, KakerMix uses vintage synths, punchy electronic drums and spacey reverb to make us feel like we’re entering another dimension. https://suno.com/album/4faa3913-892f-47ef-b2a7-63fd1848b5c9