Tootfinder

Verdächtige Onlinewetten: Spotify ändert nachträglich Song-Rangliste in den USA
Wer den meistgespielten Song auf Spotify in dieser Woche richtig getippt hat, hat einen immensen Gewinn gemacht – und womöglich von einem Betrug profitiert.

Verdächtige Onlinewetten: Spotify ändert nachträglich Song-Rangliste in den USA
Wer den meistgespielten Song auf Spotify in dieser Woche richtig getippt hat, hat einen immensen Gewinn gemacht – und womöglich von einem Betrug profitiert.

Another Ukrainian innovation:
interceptor drones are being mounted on small transport aircraft like An-28s to take down Russian attack drones
https://bsky.app/profile/mariadrutska.bsky.social/post/3mk6oatoggc2t

Maria Drutska (@mariadrutska.bsky.social)
Another Ukrainian innovation: interceptor drones are being mounted on small transport aircraft like An-28s to take down Russian attack drones. This great video shows how it works, from takeoff to interception with P1-SUN drones. I expect allies to be very interested in this.

Nice paper about why LLMs can never be fully protected against prompt injection.
#LLM #AI

Prompt Injection as Role Confusion
LLMs can't tell who's speaking. We show they identify roles by writing style, not tags, and exploit this with CoT Forgery, injecting fake reasoning that models mistake for their own thoughts.

The Book I Could Not Afford to Get Wrong
Every book I take on carries some risk, and on most of them the risk is mine alone....
https://bolesblogs.com/2026/06/02/the-book-i-could-not-afford-to-get-wrong/

The Book I Could Not Afford to Get Wrong | David Boles, Blogs
Every book I take on carries some risk, and on most of them the risk is mine alone. If I misjudge a scene or overwrite a chapter, the cost is my own time and my own name, and I can live with that. …

A detailed recap of the real-world target hacks by OpenAI's and Anthropic's models, exposing failures in AI alignment training and meaningful supervision (Zvi Mowshowitz/Don't Worry About the Vase)
https://thezvi.substack.com/p/further-developments-about-interna…

Further Developments About Internal AI Models Hacking Things
If I had a nickel for every major leading AI lab that sheepishly admitted that the model it thought was sandboxed had, during a cybersecurity evaluation with its safeguards lowered, successfully hacked outside companies, I would have two nickels.

Not re-signing Jadeveon Clowney is a mistake by the Cowboys https://insidethestar.com/not-re-signing-jadeveon-clowney-is-a-mistake-by-the-cowboys

GNStor: Design of GPU-Native High-Performance Remote All-Flash Array
Shushu Yi, Wenbo Wu, Guoci Chen, Junrong Zhu, Shengwen Liang, Mao Bo, Chenying Huan, Chen Tian, Jie Zhang
https://arxiv.org/abs/2606.04908 https://arxiv.org/pdf/2606.04908 https://arxiv.org/html/2606.04908
arXiv:2606.04908v1 Announce Type: new
Abstract: GPU has become the leading computing device for a wide range of data-intensive applications, which tightly collaborates with remote all-flash array (AFA) to accommodate ever-expanding datasets, facilitate multi-client data sharing, and guarantee fault tolerance. Although GPU is the center of computation, all I/O processes in existing GPU-AFA systems are still CPU-centric. CPU orchestrates remote I/O requests and executes a centralized AFA engine to take charge of AFA-level functionalities (e.g., access control and metadata persistence). This design disparity suffers from substantial CPU-GPU interaction overhead and I/O traffic amplification, compromising end-to-end I/O performance.
In this work, we present \emph{GNStor}, a GPU-native AFA system that enables GPU to directly access remote AFA without CPU intervention in the I/O path, thereby fully exploiting the performance of AFA. Specifically, GNStor first proposes a GPU-centric NVMe over RDMA (NoR) software stack (named \emph{GNoR}), paving a fast path for GPUs to directly initiate NoR I/O requests to SSDs within remote AFA. GNoR employs an atomic-operation-based I/O orchestration design and follows the single-instruction-multiple-thread (SIMT) execution model of GPU, fully exploiting the massive parallelism of GPU architectures. To facilitate essential AFA functionalities in a CPU-bypass I/O path, GNStor further designs \emph{deEngine}, a decentralized AFA engine that seamlessly decomposes and integrates AFA-level tasks into each SSD firmware, thereby achieving efficient AFA access at low cost. Evaluation results show that GNStor achieves 3.2$\times$ higher I/O throughput and reduces application execution time by 31.1\%, compared to state-of-the-art AFA systems.
toXiv_bot_toot

Tangerine Dream zagra w Polsce w maju przyszłego roku.
Ale jak zobaczyłem, że koncert w Warszawie odbędzie się w Progresji, podczas gdy w innych miastach w gmachach oper i filharmonii, to mam poważną zagwozdkę, czy nie pojechać gdzieś w Polskę.

Mistakes and None Thereof II ❌
错误和当中没有错误 II ❌
📷 Pentax 6x7
🎞️ Ilford HP5 Plus 400
If you like my work, Support by buying me a coffee or a roll of film from
PayPal https://www.paypal.com/paypalme/ydcdingsite
Wise

Mistakes and None Thereof III ❌
错误和当中没有错误 III ❌
📷 Nikon F4E
🎞️ Kentmere 400
If you like my work, Support by buying me a coffee or a roll of film from
PayPal https://www.paypal.com/paypalme/ydcdingsite
Wise