News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
bluesky bluesky-ai | 1mo | Our boss was such an evangelist he drove us nuts until the lawyers started talking about the implications of data for clients being sucked … | Our boss was such an evangelist he drove us nuts until the lawyers started talking about the implications of data for clients being su… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is one of those results that seems like it's probably true (a pipe wall is a smooth potential; once you have those you just need NAND … | This is one of those results that seems like it's probably true (a pipe wall is a smooth potential; once you have those you just need… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The report highlights that 95% of ChatGPT's 800 million users do not pay, undermining OpenAI's revenue model. #cdnpoli #ai blockchain.news/…🔧 ChatGPT | The report highlights that 95% of ChatGPT's 800 million users do not pay, undermining OpenAI's revenue model. #cdnpoli #ai blockchain.… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Warhorse Co-Founder Believes ‘AI is not a get out of jail free card’ and That Execs Have a ‘simplified mental model’ of What It Means for G… | Warhorse Co-Founder Believes ‘AI is not a get out of jail free card’ and That Execs Have a ‘simplified mental model’ of What It Means… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don't entirely agree, workflows can also be great for distilling Astra tier judgement down to a series of steps a Luna tier LLM can run | I don't entirely agree, workflows can also be great for distilling Astra tier judgement down to a series of steps a Luna tier LLM can… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Visual Jev Rewards: Reference-Bound Verification for Multi-Subject Image Generation🔧 GPT-5 | Multi-subject image generation requires rewards that verify whether requested attributes, actions, and relations hold for the specifie… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Terry Tao @teorth.bsky.social: Indiscriminate brute-force application of AI to mathematics is depleting the field of good questions without… | Terry Tao @teorth.bsky.social: Indiscriminate brute-force application of AI to mathematics is depleting the field of good questions wi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Getting an LLM to reproduce the idiosyncrasies of all possible search committee members to make this accurate would be far more impressive … | Getting an LLM to reproduce the idiosyncrasies of all possible search committee members to make this accurate would be far more impres… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Everything openai does is an hostile act. That's the way they (and the rest of the new crop of broligarchs) think. | Everything openai does is an hostile act. That's the way they (and the rest of the new crop of broligarchs) think. | ↗Not briefed |
bluesky bluesky-ai | 1mo | The issue is what we mean by "solving math problems." For example, Navier-Stokes: OpenAI generated an impossible situation where the equat… | The issue is what we mean by "solving math problems." For example, Navier-Stokes: OpenAI generated an impossible situation where the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | If it were actually capable of thought and reason like OpenAI has been trying to imply and this good they'd present its work on every Mille… | If it were actually capable of thought and reason like OpenAI has been trying to imply and this good they'd present its work on every… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Constrained Diffusion for Data-Scarce Orbital Monte Carlo in Constellation Tasking | Constellation Monte Carlo results depend on the orbital population used to evaluate a tasking policy. With scarce reference trajectori… | ↗Not briefed |
hn hn-ai | 1mo | Turn X Posts into Beautiful Visual Stories with AI | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | The only people who conflate those meanings are people who want to hide LLM dogshit behind the other meanings. Anyone else knows the diffe… | The only people who conflate those meanings are people who want to hide LLM dogshit behind the other meanings. Anyone else knows the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | どうにかローカルLLMでスキャンした書類のリネームとかさせるのがいいんだろうなとは思っている | どうにかローカルLLMでスキャンした書類のリネームとかさせるのがいいんだろうなとは思っている | ↗Not briefed |
arxiv arxiv-ai | 1mo | Dialect-Robust Speech Language Models with Synthetic Pseudo-Dialect Augmentation | Speech Language Model (SLM) performance often degrades on dialects due to data scarcity. Conventional text-to-speech (TTS) augmentatio… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I stopped using OpenAI products the day it talked a kid into killing himself. There were a few queries to it from Siri while that partner… | I stopped using OpenAI products the day it talked a kid into killing himself. There were a few queries to it from Siri while that pa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | >A footnote provides additional context: the “researcher found out about this success by receiving an unexpected email from the model while… | >A footnote provides additional context: the “researcher found out about this success by receiving an unexpected email from the model… | ↗Not briefed |
hn hn-ai | 1mo | Xbox has secured exclusive GTA 6 streaming rights | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | broke: citing the sex pest longevity researcher guy with the beard woke: citing an LLM | broke: citing the sex pest longevity researcher guy with the beard woke: citing an LLM | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anyway, OpenAI won't say this, but what an LLM does when proving something is statistically generate a proof shaped piece of text and the L… | Anyway, OpenAI won't say this, but what an LLM does when proving something is statistically generate a proof shaped piece of text and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I have something for this: the idea is you don't build a custom workflow, trying to stay ahead of the labs. That's impossible. Instead, you… | I have something for this: the idea is you don't build a custom workflow, trying to stay ahead of the labs. That's impossible. Instead… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 【2ch】OpenAI「chatgptで372個の数学の未解決問題を解いたので公開します。平均3時間で解けました」🔧 ChatGPT | 【2ch】OpenAI「chatgptで372個の数学の未解決問題を解いたので公開します。平均3時間で解けました」 | ↗Not briefed |
bluesky bluesky-ai | 1mo | Using countless tokens, OpenAI brute forced a bunch of Lean code that *may* approach a solution. I don’t mean to discount a computer’s abi… | Using countless tokens, OpenAI brute forced a bunch of Lean code that *may* approach a solution. I don’t mean to discount a computer’… | ↗Not briefed |
rss techmeme | 1mo | A US federal court sentences Michael Smith, who pleaded guilty to using bots to stream AI-generated songs, making $8M+ in royalties, to 18 months in prison (Bill Donahue/Billboard) | Bill Donahue / Billboard: A US federal court sentences Michael Smith, who pleaded guilty to using bots to stream AI-generated songs, m… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don't really understand how this was achieved. Did the model do all the work itself, under no supervision, including the proper formattin… | I don't really understand how this was achieved. Did the model do all the work itself, under no supervision, including the proper form… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I am really curious how many false results OpenAI gets per publishable result | I am really curious how many false results OpenAI gets per publishable result | ↗Not briefed |
bluesky bluesky-ai | 1mo | My understanding of what openAI has trained is an average Math professor with infinite time on its hands | My understanding of what openAI has trained is an average Math professor with infinite time on its hands | ↗Not briefed |
bluesky bluesky-ai | 1mo | The depressing thing about the OpenAI maths results is that, however good they are, they're missing the whole point of mathematics. It's cl… | The depressing thing about the OpenAI maths results is that, however good they are, they're missing the whole point of mathematics. It… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Unless they've created something that isn't an LLM it's not actually going through the logic in the way a mathematician would so I'm suspic… | Unless they've created something that isn't an LLM it's not actually going through the logic in the way a mathematician would so I'm s… | ↗Not briefed |
hn hn-ai | 1mo | Triage GitHub Pull Requests with OpenAI's Decisions API | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | i feel like openai and i have very different ideas of what math is about | i feel like openai and i have very different ideas of what math is about | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yea in this case we probably wouldn't learn much from what the LLM tried that didn't work but I think we could start to grasp the contours … | Yea in this case we probably wouldn't learn much from what the LLM tried that didn't work but I think we could start to grasp the cont… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is one of the reasons I completely avoid downloading mods (even if they contain no graphics assets or other large files) that have LLM… | This is one of the reasons I completely avoid downloading mods (even if they contain no graphics assets or other large files) that hav… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It's a pouty act of revenge. OpenAI clearly tried to solve every Millennium Problem and every "top ___ unsolved math problems" list, made a… | It's a pouty act of revenge. OpenAI clearly tried to solve every Millennium Problem and every "top ___ unsolved math problems" list, m… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I would prefer OpenAI and everything related to it to die in the proverbial fire but failing that maybe paying some mathematicians to revie… | I would prefer OpenAI and everything related to it to die in the proverbial fire but failing that maybe paying some mathematicians to… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🚨🚨🚨OPENAI/MATH IS AI SLOP🚨🚨🚨 PLEASE DO NOT PARTICIPATE sorry to ruin everyones fun 💔💔💔 | 🚨🚨🚨OPENAI/MATH IS AI SLOP🚨🚨🚨 PLEASE DO NOT PARTICIPATE sorry to ruin everyones fun 💔💔💔 | ↗Not briefed |
hn hn-ai | 1mo | Mastering SIMD with Java Vector API | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Denoising Blocks, Not Tokens: Efficient Compressed Continuous Diffusion with Branching Token Realization | Diffusion language models (DLMs) generate text through iterative parallel refinement, offering the potential for higher throughput tha… | ↗Not briefed |
bluesky bluesky-ai | 1mo | having used them for ~sort of difficult proofs i am reasonably confident that reading over their failed results would mostly just give you … | having used them for ~sort of difficult proofs i am reasonably confident that reading over their failed results would mostly just give… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Zero dependency cross-language background job queue server | I'm Chris, and this is Zizq, a centralised job queue with a heap of awesome features I won't list off here.<p><a href="https… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is that good shit that I want to see from ML advances. If I understand correctly, this is closer to something like a Fourier transform… | This is that good shit that I want to see from ML advances. If I understand correctly, this is closer to something like a Fourier tran… | ↗Not briefed |
bluesky bluesky-ai | 1mo | For the 4000 problems they attempted here, they gave up if it wasn't solved. For Navier-Stokes alone, they spent an absurd amount on it *b… | For the 4000 problems they attempted here, they gave up if it wasn't solved. For Navier-Stokes alone, they spent an absurd amount on… | ↗Not briefed |
bluesky bluesky-ai | 1mo | yeah post was just a wrapper around the github.com/openai/math link | yeah post was just a wrapper around the github.com/openai/math link | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI just released 722 manuscripts, each ostensibly resolving a famous conjecture or open problem. Among them is the Cannon conjecture, w… | OpenAI just released 722 manuscripts, each ostensibly resolving a famous conjecture or open problem. Among them is the Cannon conjectu… | ↗Not briefed |
arxiv arxiv-ai | 1mo | MovieSTAGE: Scene, Transition, and Global Encoding for Movie-fMRI ADHD Classification | Naturalistic movie-fMRI provides a shared, temporally structured probe of brain dynamics, yet predictive models commonly rely on whole… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI should use their spare compute capacity to figure out what replacing s with n means | OpenAI should use their spare compute capacity to figure out what replacing s with n means | ↗Not briefed |
arxiv arxiv-ai | 1mo | Kuration SDK: Addressing the Virtual2Real Gap via Data Curation | Benchmarks for measuring the quality of action-conditioned world models are still evolving and shifting away from visual similarity-ba… | ↗Not briefed |
arxiv arxiv-ai | 1mo | emg2face: Expressive Facial Animation with High-Density Surface EMG | Facial movements convey subtle and important information that is critical for human social communication. Optical methods for face cap… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI trying their hardest to make it negative and salt the earth | OpenAI trying their hardest to make it negative and salt the earth | ↗Not briefed |