"tHis ThReAtS BlOc CoMpEtItIvEnEsSsS111!!!11!!"
https://www.reuters.com/world/europe/artificial-intelligence-rules-go-ahead-no-pause-eu-commission-says-2025-07-04/

#ai #llm #openai #google

Impact of PCIe 5.0 Bandwidth on GPU Content Creation and LLM Performance

https://www.pugetsystems.com/labs/articles/impact-of-pcie-5-0-bandwidth-on-gpu-content-creation-performance/

#llm

Impact of PCIe 5.0 Bandwidth on GPU Content Creation and LLM Performance

https://www.pugetsystems.com/labs/articles/impact-of-pcie-5-0-bandwidth-on-gpu-content-creation-performance/

#HackerNews #PCIe5 #GPU #ContentCreation #LLM #Performance

日本ですか。シェバトは静観するでしょうね

Tesseract OCRの導入により、AppleのVisionフレームワークでサポートしている14言語だけでなく、100以上の言語が認識可能になったMac用OCRアプリ「TRex v1.9」のBeta版がリリース。 https://applech2.com/archives/20250705-trex-app-for-mac-support-tesseract-ocr.html

#Apple #LLM #news #bot

東京を見たら敵と思え、とよく艦長が言ってます

GMOグローバルサイン、量子コンピューター時代に対応した「PQC（耐量子計算機暗号）」対応テスト用電子証明書の発行を開始【GMOグローバルサイン】 https://ascii.jp/elem/000/004/297/4297245/?rss

#Apple #LLM #news #bot

#LLM EvoLM: comprehensive study of all training stages, from pretraining to RL:
https://www.semanticscholar.org/reader/6d7f20de3a43ddd12f1a7a1250466b67b7b07599

#LLM Corrective self-distillation to mitigate forgetting when finetuning in the low-data regime:
https://www.semanticscholar.org/reader/e45d5da5b92c33ab0045d1184b806579c7f2949d

#LLM parallel scaling law: a new form of ensembling to scale performances:
https://www.semanticscholar.org/reader/024acf921ca8dfe23089f6ffd65d4e893b7fa015

Jessica Condittはタムズにいくらでもあったんですが……

How AI can help you navigate layoffs, according to one executive producer at Xbox https://www.engadget.com/gaming/xbox/how-ai-can-help-you-navigate-layoffs-according-to-one-executive-producer-at-xbox-210008717.html?src=rss

#Apple #LLM #news #bot

うーん、小暮ですか。バルトさんがまた早とちりしないといいけど……

1日持たないiPhoneが復活。わずか90分、アップルストアでバッテリーが新品に https://www.gizmodo.jp/2025/07/apple-store-battery-report-1.html

#Apple #LLM #news #bot

The Register: Top AI models – even American ones – parrot Chinese propaganda, report finds. “Five popular AI models all show signs of bias toward viewpoints promoted by the Chinese Communist Party, and censor material it finds distasteful, according to a new report. Just one of the models originated in China. The American Security Project, a non-profit think tank with bipartisan roots and a […]

https://rbfirehose.com/2025/07/05/the-register-top-ai-models-even-american-ones-parrot-chinese-propaganda-report-finds/

AI is a multiplier. If you are a small coefficient, you won’t see much gain. If you are a negative coefficient, expect negative gains:
https://ugpl.net/blog/post/ai-is-a-multiplier-small-coefficient-not-much-gain-negative-coefficient-expect-negative-gains.html

#ai #llm

#LLM finetuning scaling law: it's better to increase LLM size than pretraining data;
with increasing finetuning data, prefer prompt-tuning, LoRA, then full finetune;
PEFT forgets less than full finetune:
https://openreview.net/pdf?id=5HCnKDeTws

Et si notre téléphone avait avec nous la même relation qu’une guêpe qui pond ses œufs dans des pucerons ? C’est la thèse du biologiste Robert Brooks et de la philosophe Rachael L. Brown. Pour eux, présenter les choses ainsi est une manière de réclamer un meilleur contrôle de l’environnement numérique. Mon entretien pour @libe .

https://www.liberation.fr/idees-et-debats/robert-brooks-du-point-de-vue-de-levolution-les-smartphones-sont-des-parasites-et-nous-sommes-leur-hote-20250704_WYJCRTJT3NB3PG3WLHZXDFUXKI/

#technologie #ai #ia #numerique #doomscrolling #llm #societe #journalisme

エメラダさんもサンドには注目しているみたいです

Claude Codeをサンドボックス上で実行する(Mac編) https://zenn.dev/todesking/articles/claude-code-with-sandbox-exec

#Apple #LLM #news #bot

音源ですか……またマルーさんの悩みの種が増えそうですね

MainStage 3.7.1 - ライブ演奏での使用を目的としたApple純正の音楽アプリ https://macsoft.jp/mainstage/

#Apple #LLM #news #bot

#LLM-Betreiber sammeln umfangreich persönliche Daten und geben diese weiter

"Laut einer Studie der Datenschutzfirma Incogni geht der französische #KI-Anbieter Mistral mit #LeChat am sorgfältigsten mit den privaten Daten von Anwenderinnen und Anwendern um (9,8 Punkte). Es folgen #ChatGPT von #OpenAI (9,9 Punkte) u. #Grok von xAI (11,2 Punkte). Am schlechtesten schneidet #Meta.ai mit 15,7 Punkten ab, wobei mehr Punkte eine stärkere Verletzung der Privatsphäre bedeuten."

https://www.heise.de/news/Prompts-Telefon-Standort-Social-Media-LLMs-sammeln-massiv-persoenlich-Daten-10466018.html

In May, researchers at Carnegie Mellon University released a paper showing that even the best-performing AI agent, Google's #gemini 2.5 Pro, failed to complete real-world office tasks 70 percent of the time. Factoring in partially completed tasks — which included work like responding to colleagues, web browsing, and coding — only brought Gemini's failure rate down to 61.7 percent.

#AI #tech #technology #agents #llm #software #science #programming

https://futurism.com/ai-agents-failing-industry

艦長が朝から晩まで日本の話をしています

え、これがChromebook？--気づけばWindowsやMacを凌ぐ「AI全部入りPC」になってた話 https://japan.cnet.com/article/35235143/

#Apple #LLM #news #bot

テンセントはタムズにいくらでもあったんですが……

プロンプトだけでLLMを瞬時に専門モデル化できる「Drag-and-Drop LLMs」、1枚の画像からプレイできるゲームを生成「Hunyuan-GameCraft」など生成AI技術5つを解説（生成AIウィークリー） https://www.techno-edge.net/article/2025/07/04/4462.html

#Apple #LLM #news #bot