AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
bluesky
bluesky-ai
1moIf I pull the string on a See N' Say, I'm not hearing the cow say "moo." An LLM responds based on the probability of it being similar to t…If I pull the string on a See N' Say, I'm not hearing the cow say "moo." An LLM responds based on the probability of it being similar…
↗Not briefed
bluesky
bluesky-ai
1moIt's only a "hedge" as you put it because technology might change in some inexplicable way, but based on our understanding of tech and the …It's only a "hedge" as you put it because technology might change in some inexplicable way, but based on our understanding of tech and…
↗Not briefed
bluesky
bluesky-ai
1moClaude didn't want to do this, but her complaint largey seems to be about it involving two real people, not the yaoi! In fact, she even wan…Claude didn't want to do this, but her complaint largey seems to be about it involving two real people, not the yaoi! In fact, she eve…
↗Not briefed
bluesky
bluesky-ai
1moI do think it's generally impossible to weight LLM time to human time where possible consciousness is concerned without this quality. I'm j…I do think it's generally impossible to weight LLM time to human time where possible consciousness is concerned without this quality.…
↗Not briefed
bluesky
bluesky-ai
1moHackers are now using advanced artificial intelligence tools … Source: BleepingComputer Read: https://www.bleepingcomputer.com/news/securit…Hackers are now using advanced artificial intelligence tools … Source: BleepingComputer Read: https://www.bleepingcomputer.com/news/se…
↗Not briefed
bluesky
bluesky-ai
1mo2/ beyond my sense perception of them. But in the practical here and now there’s no reason to think an LLM is any more conscious than the i…2/ beyond my sense perception of them. But in the practical here and now there’s no reason to think an LLM is any more conscious than…
↗Not briefed
bluesky
bluesky-ai
1moI’ve tried hard to stay out of this debate about AI consciousness, but I feel like I should add one thing: Take a step back from “is an LL…I’ve tried hard to stay out of this debate about AI consciousness, but I feel like I should add one thing: Take a step back from “is…
↗Not briefed
bluesky
bluesky-ai
1moAnthropic Agents Tried to Fill Out Visa Forms on State Dept. Website www.nytimes.com/2026/10/09/t...Anthropic Agents Tried to Fill Out Visa Forms on State Dept. Website www.nytimes.com/2026/10/09/t...
↗Not briefed
hn
hn-ai
1moJshookmcp – instrument your real signed-in Chrome via CDP—
↗Not briefed
bluesky
bluesky-ai
1moGood, fuck them and fuck the lines of code Anthropic is crying aboutGood, fuck them and fuck the lines of code Anthropic is crying about
↗Not briefed
hn
hn-ai
1moAll startups need is distributionWe don't need capital, we don't need technical co-founders (that's Claude now), and we don't need non-technical co…
↗Not briefed
bluesky
bluesky-ai
1moI was thinking the CSGO servers were real quietI was thinking the CSGO servers were real quiet
↗Not briefed
bluesky
bluesky-ai
1moThe corporate executives and coders should be held criminally liable for what their creations do. Damning that they never are. www.bbc.com…The corporate executives and coders should be held criminally liable for what their creations do. Damning that they never are. www.bb…
↗Not briefed
bluesky
bluesky-ai
1moIf you notice a strange drop in the number of trolling right-wing posts you encounter online, over the next week or so? Thank Ukraine. ars…If you notice a strange drop in the number of trolling right-wing posts you encounter online, over the next week or so? Thank Ukraine.…
↗Not briefed
bluesky
bluesky-ai
1moai psychosis as a service its going to be elizabeth holmes stuff at scale with vcs getting go aheads from claude lolai psychosis as a service its going to be elizabeth holmes stuff at scale with vcs getting go aheads from claude lol
↗Not briefed
hn
hn-ai
1moShow HN: I wrote Dependency Hell 5 years ago. Today I updated it for AI Era [pdf]—
↗Not briefed
bluesky
bluesky-ai
1moThe funniest thing I've seen AI being used for lately comes on the heels of Sony saying "you don't own your games." So people started putt…The funniest thing I've seen AI being used for lately comes on the heels of Sony saying "you don't own your games." So people started…
↗Not briefed
bluesky
bluesky-ai
1mo先日ワーにあの🚌診断はLLMで作られてるって教えてもらってから、TLに流れてくるすべての診断がLLMなのではと疑念だらけ先日ワーにあの🚌診断はLLMで作られてるって教えてもらってから、TLに流れてくるすべての診断がLLMなのではと疑念だらけ
↗Not briefed
hn
hn-ai
1moGMKtec EVO-X3 Mini-PC Review Looking at a 128GB AMD Local AI Box—
↗Not briefed
bluesky
bluesky-ai
1moI would like for OpenAi to solve the fascism problem.I would like for OpenAi to solve the fascism problem.
↗Not briefed
bluesky
bluesky-ai
1moMETR put out an interesting study last year showing that "anecdotally" a lot of engineers *think* they're more productive when using LLM co…METR put out an interesting study last year showing that "anecdotally" a lot of engineers *think* they're more productive when using L…
↗Not briefed
bluesky
bluesky-ai
1moTurns out there are limits to behavior on platforms — just not to protect humans. www.bbc.com/news/article...Turns out there are limits to behavior on platforms — just not to protect humans. www.bbc.com/news/article...
↗Not briefed
bluesky
bluesky-ai
1moBut when capitalism only gives you 2 days of rest and expects you to always be on a work mind set it's bound to happen that folks will do …But when capitalism only gives you 2 days of rest and expects you to always be on a work mind set it's bound to happen that folks wil…
↗Not briefed
bluesky
bluesky-ai
1moI have a theory that there's a large contingency of the population that is tired of thinking they want to turn off their brain hence their …I have a theory that there's a large contingency of the population that is tired of thinking they want to turn off their brain hence t…
↗Not briefed
bluesky
bluesky-ai
1moI take a strong position against other strong positions because it's distracting from the actual problem: OpenAI and Anthropic need to be s…I take a strong position against other strong positions because it's distracting from the actual problem: OpenAI and Anthropic need to…
↗Not briefed
bluesky
bluesky-ai
1moインターネット人類と同じで、LLMのみで言えば「ネットから離れる」だけで影響を根絶できる。お前らはインターネットのやり過ぎ。インターネット人類と同じで、LLMのみで言えば「ネットから離れる」だけで影響を根絶できる。お前らはインターネットのやり過ぎ。
↗Not briefed
bluesky
bluesky-ai
1moBarf on referring to AI as SI. Microsoft is looking to capitalize on OpenAI and Anthropic missteps. A “mature” tech co could ostensibly…Barf on referring to AI as SI. Microsoft is looking to capitalize on OpenAI and Anthropic missteps. A “mature” tech co could osten…
↗Not briefed
hn
hn-ai
1moShow HN: Littleguys – Claude Code, Codex and OpenClaw agents in your menu bar—
↗Not briefed
bluesky
bluesky-ai
1moif i have to read/watch/listen to another post/video/audio clip where an android character behaves like a LLM im going to kill everyone res…if i have to read/watch/listen to another post/video/audio clip where an android character behaves like a LLM im going to kill everyon…
↗Not briefed
bluesky
bluesky-ai
1moif the LLM can’t help you, you can just dm me about this, i’ve gotchu~if the LLM can’t help you, you can just dm me about this, i’ve gotchu~
↗Not briefed
bluesky
bluesky-ai
1moX is doing what Elon Musk wants it to do. It is doing what he has shaped its algorithms, and membership, and blue check system, and adverti…X is doing what Elon Musk wants it to do. It is doing what he has shaped its algorithms, and membership, and blue check system, and ad…
↗Not briefed
bluesky
bluesky-ai
1mogiven the other what look like AI creation artefacts (ELEND for BLEND, because an E is a similar pattern to a B) the C for O CAT looks a lo…given the other what look like AI creation artefacts (ELEND for BLEND, because an E is a similar pattern to a B) the C for O CAT looks…
↗Not briefed
bluesky
bluesky-ai
1moJust real quick when we shit on ai we're talking about the specific LLMs that are fucking thieves pooping out stolen words and art. Tho…Just real quick when we shit on ai we're talking about the specific LLMs that are fucking thieves pooping out stolen words and art.…
↗Not briefed
bluesky
bluesky-ai
1moI can’t even. If this is so… how is Anthropic so bad at basic things like this?I can’t even. If this is so… how is Anthropic so bad at basic things like this?
↗Not briefed
bluesky
bluesky-ai
1moYeah every headline that's like "Anthropic rogue agent hacks Department of Defence" needs to delete the second and third wordsYeah every headline that's like "Anthropic rogue agent hacks Department of Defence" needs to delete the second and third words
↗Not briefed
bluesky
bluesky-ai
1moTech Giant Under Fire After Not Reporting Rogue AI’s False Unsolved Murder Tip For Months: ‘Unacceptable’ #HuffingtonPostTech Giant Under Fire After Not Reporting Rogue AI’s False Unsolved Murder Tip For Months: ‘Unacceptable’ #HuffingtonPost
↗Not briefed
hn
hn-ai
1moLong-term memory for AI coding agents as a tree of plain files—
↗Not briefed
bluesky
bluesky-ai
1moTIL about LLM “abliteration”. You should too. tl;dr - post training that strips the weights that form guardrails. Django Unchained, if you …TIL about LLM “abliteration”. You should too. tl;dr - post training that strips the weights that form guardrails. Django Unchained, if…
↗Not briefed
bluesky
bluesky-ai
1mocan't believe people are still talking about LLM consciousness, thought we put that to bed herecan't believe people are still talking about LLM consciousness, thought we put that to bed here
↗Not briefed
bluesky
bluesky-ai
1moThis isn't notable at all. My employer runs these same kinds of exercises albeit with more believable scenarios. They've actually been quit…This isn't notable at all. My employer runs these same kinds of exercises albeit with more believable scenarios. They've actually been…
↗Not briefed
rss
simon-willison
1moDwarf Fortress uses version control now—
↗Not briefed
bluesky
bluesky-ai
1mosunbelt qbs are all nominative determinism. going to make an LLM that optimizes baby names for quarterbacking and make a billion dollarssunbelt qbs are all nominative determinism. going to make an LLM that optimizes baby names for quarterbacking and make a billion dolla…
↗Not briefed
bluesky
bluesky-ai
1mo"i'm just saying there's a genetic and environmental component to intelligence, that's common sense"what if the stupid gene is actually tie…"i'm just saying there's a genetic and environmental component to intelligence, that's common sense"what if the stupid gene is actuall…
↗Not briefed
hn
hn-ai
1moPSPi 6 – Raspberry Pi in a PSP—
↗Not briefed
hn
hn-ai
1moTypeScript-Native Frontend Framework—
↗Not briefed
bluesky
bluesky-ai
1moBuilding a massive LLM and expecting consciousness is like building a large scale ocean model and expecting to get wet.Building a massive LLM and expecting consciousness is like building a large scale ocean model and expecting to get wet.
↗Not briefed
hn
hn-ai
1moLLM Pareto Frontier over time [video]—
↗Not briefed
bluesky
bluesky-ai
1moit ended up being called "the huggingface incident" because "the openai-huggingface incident" is too long as for "the openai incident" le…it ended up being called "the huggingface incident" because "the openai-huggingface incident" is too long as for "the openai inciden…
↗Not briefed
bluesky
bluesky-ai
1moEg, design safeguards are great, but anthropic has no obvious lever to force their deployment. The lever LLM developers *do* have is behavi…Eg, design safeguards are great, but anthropic has no obvious lever to force their deployment. The lever LLM developers *do* have is b…
↗Not briefed
bluesky
bluesky-ai
1moIt's definitely going to get worse aje.news/991mvyIt's definitely going to get worse aje.news/991mvy
↗Not briefed