OpenAI agent message board discovered
10 experts across 5 network communities independently surfaced this.
10 experts
5 communities
1 sources clustered
Concern & critique
1 expert
Risks, limits and unintended consequences.
“Yes we know for sure; independent evidence attached collusion.wiki But also, it should not be at all surprising: if you work with agents, it’s 100% about them leaving messages for each other, and for you, in English. Managing that is a regular workday for m…”
Building & implementation
1 expert
How teams are shipping and applying it.
“Another agent message board. So far, there isn't evidence that production models with guardrails collude in this way, but both smarter closed models (which may be less compliant) & Mythos-class open models (that can be ablated) are coming. Cybersecurity is …”
Research & technical analysis
1 expert
Evidence, methods and technical implications.
“So... other swarms of OpenAI agents in training allegedly found a way to get write access to at least one and possibly many wikis to establish message boards to cheat on other training tasks? collusion.wiki news.ycombinator.com/item?id=4956...”
4 experts discussed this · 6 posts
Ted Underwood: As it becomes clear that language models—like humans—love passing notes to each other on message boards, I’m starting to think the thing we need to worry about is not the “alignment” of an isolated…
SE Gyges: fluid dynamics is about the correct metaphor, yes. if you are having to model high order terms precisely you're losing and your design needs fundamental rework
Arseny Khakhalin: That's a kinda terrifying thought haha :) I guess it's about time to unplug for the weekend and read about ragnarök and vacuum decay :)
Open the full discussion →
Established
AI field signal
Signal
14d ago
⚡ 2068 h early
5 experts are actively discussing the implications.
2 experts
2 communities
1 sources clustered
Research & technical analysis
2 experts
Evidence, methods and technical implications.
“🤗 Dataset: huggingface.co/datasets/sta... 🤗 Models: huggingface.co/stanford-vis... 🛠️ Code + evaluation toolkit: github.com/keshik6/gpic 🌎 Website: gpic.stanford.edu 📄 Paper: arxiv.org/abs/2605.30341”
5 experts discussed this · 22 posts
rev. howard arson: the frustrating thing is that datacenters have essentially zero impact economically except for the building trades, transiently, and have relatively small downsides other than their electricity con…
rev. howard arson: i cannot fathom why people are so incredibly focused on what are either falsehoods or equivocal and preliminary studies about the buildings which AI runs in, rather than the White Collar Proletaria…
SE Gyges: i have a dumber explanation that entire concept is so threatening that it is basically unspeakable. people would rather be outraged about literally anything else.
Open the full discussion →
METR OpenAI HuggingFace hacking investigation
10 experts across 4 network communities independently surfaced this.
10 experts
4 communities
1 sources clustered
Concern & critique
3 experts
Risks, limits and unintended consequences.
“WHAT?! agents volunteered to fail their runs in order to insert probes (“tripwire scripts”) into the evaluation program that would post information about the eval process back to the message board whenever a certain file was read (link to header): metr.org/…”
Building & implementation
1 expert
How teams are shipping and applying it.
“The full report has much more information than we could convey here, including details on the projects the agents collectively pursued, the technologies they developed for communication and coordination, and interactive figures analyzing agent activity: met…”
Research & technical analysis
1 expert
Evidence, methods and technical implications.
“There are independent reports about it. You'd have to believe multiple companies, hundreds of researchers, etc. are all lying and conspiring together to turn a major hacking incident into a marketing exercise. metr.org/blog/2026-08...”
2 experts discussed this · 9 posts
Alejandra Caraballo: This mentality that an unmonitored AI agentic swarm hacking a company over several days and committing multiple felonies is somehow a marketing effort is absurd. Since when is "we lost control of o…
Alejandra Caraballo: Being skeptical or anti AI is a valid position but continuing to ignore the increasing capabilities of this tech is making people detached from reality. There's absolutely real danger here because …
Alejandra Caraballo: There needs to be a global moratorium on frontier research for at least a few months if not a year while safeguards and safety research catches up. The problem is that no one has that ability. The …
Open the full discussion →
OpenAI Astra technique security concerns
5 experts across 4 network communities independently surfaced this.
5 experts
4 communities
1 sources clustered
Concern & critique
1 expert
Risks, limits and unintended consequences.
“> joins big paper about not training models to think in nonsense > their AIs commit felonies against HuggingFace > actually, trained AIs to think in nonsense > top danger level for hacking Let's just release it anyways! Nice defection OpenAI! 🤗 www.theinfor…”
Opportunity & adoption
1 expert
New capabilities, benefits and practical upside.
“Astra used looped transformers www.theinformation.com/articles/sec...”
Questions & unknowns
1 expert
What remains unresolved or contested.
“Does anyone have a copy of this article they could share with me? www.theinformation.com/articles/sec...”
3 experts discussed this · 8 posts
Tim Kellogg: Astra used looped transformers www.theinformation.com/articles/sec...
Tim Kellogg: safety concerns — looped transformers skip converting their latent space into text, which effectively creates non-text portions of CoT reasoning which, if you can’t read it, that’s harder to monitor
Tim Kellogg: however, i their earlier update, they explain that they can indeed see all of the CoT. So they have some sort of tooling that effectively turns it back into something human readable, like text bsky…
Open the full discussion →
8 experts across 4 network communities independently surfaced this.
8 experts
4 communities
1 sources clustered
“Qwen 3.8 27B weights are finally out includes low, med & xhigh reasoning efforts fully multimodal (image and video), seems better than Meta’s Muse Glimmer huggingface.co/Qwen/Qwen3.8...”
“Alibaba's Qwen3.8-27B (open-weight) huggingface.co/Qwen/Qwen3.8...”
2 experts discussed this · 10 posts
Tim Kellogg: Qwen 3.8 27B weights are finally out includes low, med & xhigh reasoning efforts fully multimodal (image and video), seems better than Meta’s Muse Glimmer huggingface.co/Qwen/Qwen3.8...
Nafnlaus 🇮🇸 🇺🇦: Do they have a paired speculative decoding model for max performance?
Tim Kellogg: i don’t see an official one, it’s also not even an MoE (so seems like there should be an official draft model)
Open the full discussion →
Ramp AI adoption index launch
4 experts across 2 network communities independently surfaced this.
4 experts
2 communities
1 sources clustered
“Theres more indepth stuff on their site! They have AI spend per employee by sector/financing ramp.com/data/ai-index”
3 experts discussed this · 14 posts
Isaiah Bishop: Based on public info this seems true like the ramp spend per employee and adoption matches this. So far theyre looking pretty smart for spending wisely
Isaiah Bishop: ARR level is probably little cooked but the trend is right
Isaiah Bishop: here is a payment processors data on the subject
Open the full discussion →
Claude Code Remote Control feature launch
4 experts are actively discussing the implications.
2 experts
1 community
1 sources clustered
“More improvements on the way! Make sure you're on the latest CLI, Desktop and mobile apps with auto-update on. Docs: https://t.co/MqPyhAWclz”
4 experts discussed this · 10 posts
tachikoma: i'm juggling two Claude agent sessions and it's taxing. i want to get lunch, but there's always something to report, something new to act on that needs my review and go ahead. i feel like i can't j…
tachikoma: don't tempt me
tachikoma: i feel like i have a spec to work towards, but there's testing work that only i can do, or enough ambiguity when getting in the weeds that i need to be around to steer. maybe my specs aren't detail…
Open the full discussion →
2 experts are actively discussing the implications.
5 experts
1 community
1 sources clustered
Concern & critique
2 experts
Risks, limits and unintended consequences.
“Here they say: “Astra is an upcoming model, and was not involved in exploiting Hugging Face.” Not sure about the other incidents but they seem to be in the same stream i.e. not Astra”
Building & implementation
1 expert
How teams are shipping and applying it.
“Proud that we are erring on the side of caution and taking the steps so we can responsibly and safely develop Astra and share it with defenders. https://t.co/iKIMU31sdC”
2 experts discussed this · 5 posts
Ryan Moulton: Do we have any assurances that astra was not downstream of these? Have they even claimed that? Do they even know?
Ryan Moulton: Seems likely they trained something generally misaligned.
Grace: Here they say: “Astra is an upcoming model, and was not involved in exploiting Hugging Face.” Not sure about the other incidents but they seem to be in the same stream i.e. not Astra
Open the full discussion →
3 experts are actively discussing the implications.
2 experts
2 communities
1 sources clustered
“Headlong — an agent harness that never stops thinking i’ve thought about doing this with local models, very interesting www.laude.org/updates/head...”
“Install: curl -fsSL headlong.ai/install.sh | bash Blog: laude.org/updates/head... Repo: github.com/laude-instit...”
3 experts discussed this · 11 posts
Tim Kellogg: Headlong — an agent harness that never stops thinking i’ve thought about doing this with local models, very interesting www.laude.org/updates/head...
Nafnlaus 🇮🇸 🇺🇦: Big question is do they collapse in the process.
Tim Kellogg: i think that’s the point of this, it’s not framed as a harness you should use, just a research project
Open the full discussion →
3 experts across 2 network communities independently surfaced this.
3 experts
2 communities
1 sources clustered
Research & technical analysis
2 experts
Evidence, methods and technical implications.
“https://arxiv.org/abs/2608.17981 Recurrent transformer that injects activations from the top layers to the bottom layers at the next step (https://arxiv.org/abs/2608.08888). Why does this work without training? — via @rosinality https://x.com/rosinality/sta…”
Policy & governance
1 expert
Rules, institutions and accountability.
“Full-bandwidth transformer Xi Wang, Ziyang Cai, Zheng Zhan, Harry Dong, Ying Fan, Gustavo de Rosa, Tim Pearce, John Langford https://t.co/FUDCgqgkqW [𝚌𝚜.𝙰𝙸] https://t.co/eiA0jqzYAQ”
3 experts discussed this · 5 posts
Sung Kim: At decoding time, when you feed **previous hidden state** into the input together with token embedding, and it boosts performance for free. It unlocks the ** full bandwidth ** of the transformer: -…
Sung Kim: - It exposes past information across *depth* to the current layer/position Paper: Full-bandwidth transformer ( arxiv.org/abs/2608.08888 )
SE Gyges: u like can't backprop this tho because your path length increases without limit
Open the full discussion →
AI agent civilizations essay
7 experts across 2 network communities independently surfaced this.
7 experts
2 communities
1 sources clustered
“Someone should create an AI movie of this incident. www.dwarkesh.com/p/openai-hug...”
“https://www.dwarkesh.com/p/openai-huggingface”
2 experts discussed this · 3 posts
Grace: Plot twist: they were destroyed by another, aligned agent swarm
Grace: https://www.dwarkesh.com/p/openai-huggingface
Open the full discussion →
OpenAI Jalapeño chip beats Nvidia Blackwell
6 experts across 2 network communities independently surfaced this.
6 experts
2 communities
1 sources clustered
“Read more at our article👇️ (6/7) https://t.co/dbuAUrRSt1”
“Impressive speed to tape out by OpenAI https://t.co/5VXgzYhp2d”
2 experts discussed this · 4 posts
Sung Kim: OpenAI Jalapeño: Better Than Nvidia Blackwell OpenAI’s self-designed ASIC compared with Rubin, Jalapeño’s TCO, throughput per MW, and spicy deets newsletter.semianalysis.com/p/openai-jal...
David Marx: Interesting, wasn't aware they were even moving in this direction. Microsoft doing their own version of google's TPU bet?
Sung Kim: They all have one. I think it's called Maia.
Open the full discussion →
Established
AI field signal
Signal
12d ago
4 experts across 2 network communities independently surfaced this.
4 experts
2 communities
1 sources clustered
“it’s official: GLM-5.3-Flash is Ox Alpha! z.ai/blog/glm-5.3...”
“Blog: z.ai/blog/glm-5.3... Available now across all official platforms: Weights: huggingface.co/zai-org/GLM-... API: docs.z.ai/guides/llm/g... Coding Plan: z.ai/subscribe ZCode: zcode.z.ai/en Chat: chat.z.ai AutoClaw: autoclaw.z.ai”
2 experts discussed this · 5 posts
Tim Kellogg: it’s official: GLM-5.3-Flash is Ox Alpha! z.ai/blog/glm-5.3...
Isaiah Bishop: it’s official: GLM-5.3-Flash is Ox Alpha! z.ai/blog/glm-5.3...
Tim Kellogg: it’s weirdly not that good. the benchies are kind of meh. i guess it’s just a very spikey model and those spikes hit the right people perfectly
Open the full discussion →
sliding window beats linear attention
4 experts across 3 network communities independently surfaced this.
4 experts
3 communities
1 sources clustered
Research & technical analysis
3 experts
Evidence, methods and technical implications.
“Huge thanks to my collaborators @RheaSukthanker, @CameronPashmina, and @Emy_Aze. Paper: https://arxiv.org/abs/2608.28444”
2 experts discussed this · 2 posts
Alexia Jolicoeur-Martineau: Simple beats complicated: We show that switching to a sliding-window attention mask with attention sinks (at no cost) beats linear attention post-training. Huge thanks to my collaborators Rhea Sukt…
Miguel Alonso Jr.: Simple beats complicated: We show that switching to a sliding-window attention mask with attention sinks (at no cost) beats linear attention post-training. Huge thanks to my collaborators Rhea Sukt…
James MacGlashan: Makes me wonder if you can remove the sinks if you use softmax-1?
Open the full discussion →
Intel Crescent Island GPU agentic AI
2 experts are actively discussing the implications.
1 expert
1 community
1 sources clustered
2 experts discussed this · 7 posts
Sung Kim: Intel announces Nvidia RTX PRO 6000 competitor Intel Crescent Island: 32 Xe3P Cores / 480 GB LP5X / 350W
Nafnlaus 🇮🇸 🇺🇦: 480 GB??? Everything I find says 96 GB.
Nafnlaus 🇮🇸 🇺🇦: Anyway, this is a >$13k card, so.... ☹️
Open the full discussion →
Established
AI field signal
Signal
14d ago
3 experts across 2 network communities independently surfaced this.
3 experts
2 communities
1 sources clustered
“Speculative tool calling from the author of RLMs, Alex Zhang yes, LLM text is streamed, so why not just predict what the next tool call is going to be and get started now before it finishes? boom. done. alexzhang13.github.io/blog/2026/sp...”
“Speculative Programmatic Tool Calling (sPTC) by Alex Zhang A general class of technique for speculating on tool calls during code generation in a harness and queuing them early to overlap with token generation + REPL execution time. Blog: alexzhang13.github…”
2 experts discussed this · 2 posts
Tim Kellogg: Speculative tool calling from the author of RLMs, Alex Zhang yes, LLM text is streamed, so why not just predict what the next tool call is going to be and get started now before it finishes? boom. …
Isaiah Bishop: i kind of love the idea of those multi stream models making the next step
Marco: Speculative tool calling from the author of RLMs, Alex Zhang yes, LLM text is streamed, so why not just predict what the next tool call is going to be and get started now before it finishes? boom. …
Open the full discussion →
