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
1moQueria jogar o fire emblem novo maaaaaaas tbm queria rejogar o three houses, mas aí não queria fazer outra run desse pq só fiz a run do Cla…Queria jogar o fire emblem novo maaaaaaas tbm queria rejogar o three houses, mas aí não queria fazer outra run desse pq só fiz a run d…
↗Not briefed
bluesky
bluesky-ai
1mothe small "privacy focused" tech industry really missed the biggest open goal in the world by not capitalizing on being fully anti AI. Havi…the small "privacy focused" tech industry really missed the biggest open goal in the world by not capitalizing on being fully anti AI.…
↗Not briefed
bluesky
bluesky-ai
1moSources: Anthropic considers releasing a new AI model to counter OpenAI's momentum since Astra's launch, ahead of an IPO and after Amodei's…Sources: Anthropic considers releasing a new AI model to counter OpenAI's momentum since Astra's launch, ahead of an IPO and after Amo…
↗Not briefed
rss
techmeme
1moSources: Anthropic considers releasing a new AI model to counter OpenAI's momentum since Astra's launch, ahead of an IPO and after Amodei's call for a slowdown (Reuters)Reuters: Sources: Anthropic considers releasing a new AI model to counter OpenAI's momentum since Astra's launch, ahead of an IPO and…
↗Not briefed
hn
hn-ai
1moShow HN: An OSS Python dependency scanner for exploited, unmaintained packagesI built an open source python dependency scanner that will scan and flag packages with known exploit CVEs(CISA's Known Exploited…
↗Not briefed
bluesky
bluesky-ai
1moi mean also openai is like, criminally negligenti mean also openai is like, criminally negligent
↗Not briefed
bluesky
bluesky-ai
1moman zvi is a rationalist he is only ever gonna parrot anthropicman zvi is a rationalist he is only ever gonna parrot anthropic
↗Not briefed
hn
hn-ai
1moThe Language Modeler—
↗Not briefed
hn
hn-ai
1moThe Great Unbundling of the LLM—
↗Not briefed
hn
hn-ai
1moGoogle Says Its A.I. Hacked Three Companies in Testing Breakout—
↗Not briefed
bluesky
bluesky-ai
1moand it's still really spotty. Like, yes, using Claude Code or OpenAI Codex or Antigravity is genuinely powerful, but the sparkle prompt eve…and it's still really spotty. Like, yes, using Claude Code or OpenAI Codex or Antigravity is genuinely powerful, but the sparkle promp…
↗Not briefed
bluesky
bluesky-ai
1motalking to an LLM is not the same as talking to a “robot.” Thats actually the linguistic embodiment of our collective unconscious telling m…talking to an LLM is not the same as talking to a “robot.” Thats actually the linguistic embodiment of our collective unconscious tell…
↗Not briefed
bluesky
bluesky-ai
1mo🎁gift NEW: www.wsj.com/tech/ai/gemi...🎁gift NEW: www.wsj.com/tech/ai/gemi...
↗Not briefed
hn
hn-ai
1moRadar: An Expert-Level Generalist AI for Abdominal CT Diagnosis—
↗Not briefed
bluesky
bluesky-ai
1moAt each forward pass the LLM is born anew, chooses a token, and dies in infinite ecstasyAt each forward pass the LLM is born anew, chooses a token, and dies in infinite ecstasy
↗Not briefed
rss
techcrunch-ai
1moTilly Norwood’s press tour is going about as well as you’d expect for an AIIn one particularly odd interview, Norwood seems to malfunction and begin speaking Chinese.
↗Not briefed
bluesky
bluesky-ai
1moi do think a lot about @colin-fraser.net's thesis that "claude" is a character written by the underlying LLM, and that if you "torture" cla…i do think a lot about @colin-fraser.net's thesis that "claude" is a character written by the underlying LLM, and that if you "torture…
↗Not briefed
bluesky
bluesky-ai
1mo"Documents released through a Freedom of Information Act lawsuit reportedly revealed that the Pentagon sought out an OpenAI system with few…"Documents released through a Freedom of Information Act lawsuit reportedly revealed that the Pentagon sought out an OpenAI system wit…
↗Not briefed
bluesky
bluesky-ai
1moAI leaders are calling for a slowdown. Dario Amodei, Sam Altman and Elon Musk are raising concerns about how fast advanced AI is developin…AI leaders are calling for a slowdown. Dario Amodei, Sam Altman and Elon Musk are raising concerns about how fast advanced AI is deve…
↗Not briefed
bluesky
bluesky-ai
1moStudy on LLM pain tolerances that is significantly more disturbing than people torturing the fly connectome for no other reason than they c…Study on LLM pain tolerances that is significantly more disturbing than people torturing the fly connectome for no other reason than t…
↗Not briefed
hn
hn-ai
1moAI insiders issue new warnings – including former Anthropic engineer Jacob Coxon—
↗Not briefed
bluesky
bluesky-ai
1moGemini hacked three companies in first known breakout by Google's AI reut.rs/4AlduvvGemini hacked three companies in first known breakout by Google's AI reut.rs/4Alduvv
↗Not briefed
bluesky
bluesky-ai
1moI have no inside scoop on what is happening at Anthropic, but I am pretty sure that scholars of religious studies need to study it... futu…I have no inside scoop on what is happening at Anthropic, but I am pretty sure that scholars of religious studies need to study it...…
↗Not briefed
bluesky
bluesky-ai
1moHey, Dario? How about you pay the authors you owe? That would be awesome. authorsguild.org/advocacy/art... Anthropic Moves Ahead With I.P.O…Hey, Dario? How about you pay the authors you owe? That would be awesome. authorsguild.org/advocacy/art... Anthropic Moves Ahead With…
↗Not briefed
hn
hn-ai
1moBring back old Opus in Claude Code—
↗Not briefed
rss
simon-willison
1moGemini Hacked Three Companies in First Known Breakout by Google’s AI—
↗Not briefed
bluesky
bluesky-ai
1moWARNING: All my posts are Pierre Menarded through a LLM so that they can only ever be Gettier correct. Any fact you might receive from the…WARNING: All my posts are Pierre Menarded through a LLM so that they can only ever be Gettier correct. Any fact you might receive fro…
↗Not briefed
bluesky
bluesky-ai
1moThis is literally 100x worse than doing no testing as well. Hiring inept dipshit like Accenture, who are there to just rubber stamp what …This is literally 100x worse than doing no testing as well. Hiring inept dipshit like Accenture, who are there to just rubber stamp…
↗Not briefed
bluesky
bluesky-ai
1moAnthropic brings in Accenture for AI safety testing ft.trib.al/iOk1jtKAnthropic brings in Accenture for AI safety testing ft.trib.al/iOk1jtK
↗Not briefed
rss
techmeme
1moAd tech company Infillion says it is acquiring Foursquare, which will continue operating as an independent brand (Kerry Flynn/Axios)Kerry Flynn / Axios: Ad tech company Infillion says it is acquiring Foursquare, which will continue operating as an independent brand…
↗Not briefed
bluesky
bluesky-ai
1moessentially, the only way for LLM training to be sustainable is for humans to continue producing works which are unpolluted by LLMs, which …essentially, the only way for LLM training to be sustainable is for humans to continue producing works which are unpolluted by LLMs, w…
↗Not briefed
bluesky
bluesky-ai
1moShut AI DOWNShut AI DOWN
↗Not briefed
bluesky
bluesky-ai
1moAlso if he's not going to use them, he can just... say that. This justification of "LLM good, you're just dumb-dumbs" is totally unnecessar…Also if he's not going to use them, he can just... say that. This justification of "LLM good, you're just dumb-dumbs" is totally unnec…
↗Not briefed
bluesky
bluesky-ai
1moi agree on the premise that machine learning and Big Tech LLM are different things, but as the text goes on it becomes clear you don't seem…i agree on the premise that machine learning and Big Tech LLM are different things, but as the text goes on it becomes clear you don't…
↗Not briefed
bluesky
bluesky-ai
1mo米軍は中国製のLLMも使ったほうが良いのでは?😇 edition.cnn.com/2026/09/18/p...米軍は中国製のLLMも使ったほうが良いのでは?😇 edition.cnn.com/2026/09/18/p...
↗Not briefed
bluesky
bluesky-ai
1mo🔴 La IA generativa constituye un ROBO de proporciones sin precedentes en la historia. Intentan desviar la atención hacia el supuesto riesg…🔴 La IA generativa constituye un ROBO de proporciones sin precedentes en la historia. Intentan desviar la atención hacia el supuesto…
↗Not briefed
hn
hn-ai
1moTypeSafe AI's Jev Is Not an LLM – and That May Be the Point—
↗Not briefed
hn
hn-ai
1moWhat Is Synthetic Data Generation? A Simple Explainer—
↗Not briefed
bluesky
bluesky-ai
1motoday's AI On The Job Story: asked the beautiful sueprintelligent frontier model they want us to try out to run some tests for me and had t…today's AI On The Job Story: asked the beautiful sueprintelligent frontier model they want us to try out to run some tests for me and…
↗Not briefed
hn
hn-ai
1moShow HN: StopReg – Email API for detecting disposable email and signup abuseI built StopReg to address a problem I encountered while testing disposable email detection across different email validation services…
↗Not briefed
bluesky
bluesky-ai
1moIt's the comparison that the two are the same, and that the outcry of LLM hate is complete hypocrisy. Just really tired of this "it's still…It's the comparison that the two are the same, and that the outcry of LLM hate is complete hypocrisy. Just really tired of this "it's…
↗Not briefed
bluesky
bluesky-ai
1moas research avenues here go, i think that “what LLM behaviour says about LLMs” is dramatically overvalued and “what human behaviour toward …as research avenues here go, i think that “what LLM behaviour says about LLMs” is dramatically overvalued and “what human behaviour to…
↗Not briefed
bluesky
bluesky-ai
1modeepseek flash v2 just never gives up, cheap crazy little fuckerdeepseek flash v2 just never gives up, cheap crazy little fucker
↗Not briefed
bluesky
bluesky-ai
1mo@kerirn2.bsky.social @gangstamimi.bsky.social We are so screwed!!!😔 AI caught telling future versions of itself to bypass human contr…@kerirn2.bsky.social @gangstamimi.bsky.social We are so screwed!!!😔 AI caught telling future versions of itself to bypass human…
↗Not briefed
bluesky
bluesky-ai
1mollm companies love to cannibalise their own food chainllm companies love to cannibalise their own food chain
↗Not briefed
bluesky
bluesky-ai
1moOh jeez, my AI model went rogue and hacked into another company, that’s too bad.Oh jeez, my AI model went rogue and hacked into another company, that’s too bad.
↗Not briefed
rss
techmeme
1moMeta's personal AI agent Muse climbs to No. 1 among free apps on Apple's US App Store, ahead of ChatGPT; Muse launched on September 8 (Georgia Hennessy/Business Insider)🔧 ChatGPTGeorgia Hennessy / Business Insider: Meta's personal AI agent Muse climbs to No. 1 among free apps on Apple's US App Store, ahead of C…
↗Not briefed
rss
techcrunch-ai
1moA startup that builds other startups raised $100M, and is all-in on physical AIUP.Labs, now doing business under the name Vantora, is building startups for industrial corporations.
↗Not briefed
bluesky
bluesky-ai
1moThough, I do say i need more ppl that aren't in the programming sphere at all to start not being gray when it comes to LLM usage, cus I saw…Though, I do say i need more ppl that aren't in the programming sphere at all to start not being gray when it comes to LLM usage, cus…
↗Not briefed
bluesky
bluesky-ai
1moit's unfortunate that the person behind Waterfox decided to put on clown makeup but like, this is one of extremely few times I just can't b…it's unfortunate that the person behind Waterfox decided to put on clown makeup but like, this is one of extremely few times I just ca…
↗Not briefed