📌 TOPINDIATOURS Eksklusif ai: 4Chan-Obsessed ICE Shooter Was Gaming on Steam Right

Right before he shot up a US Immigration and Customs Enforcement (ICE) office in a murderous rampage this week, Joshua Jahn of Texas was playing first-person shooter games on Steam — a startling collision point where real life and the online world seemed to bleed into each other.

Jahn’s astonishing gaming habit — more than 17,000 hours on Steam in total — was revealed in a Substack piece by investigative journalist Ken Klippenstein, who also learned from Jahn’s friends that he was a devoted habitué of 4chan, the longstanding imageboard that’s been linked to other mass shooters.

A Steam screenshot obtained by Klippenstein shows that Jahn was playing first-person shooter games Team Fortress 2 and Left 4 Dead 2 in the wee morning hours of Wednesday before he attacked the ICE office in Dallas at around 6:30 am.

Jahn found an elevated position on a neighboring building and started shooting indiscriminately, killing two immigrant detainees and leaving one other person injured. None of the victims were ICE agents. Jahn later killed himself at the scene.

This newly revealed information linking Jahn to first-person-shooter video games and 4chan will surely add more fuel to the ongoing debate on what is driving people, mostly young men, to engage in violent acts against the public or political targets such as right wing influencer Charlie Kirk, who was assassinated earlier this month at the age of 31, and President Donald Trump, who had his own brush with death last year during a campaign rally in Pennsylvania.

The ICE shooting, along with Kirk’s killing by a suspect who’s another avid gamer, 22-year-old Tyler Robinson, has revived in the public the controversial notion that games are a harmful influence — an idea that earned currency in the days after the 1999 Columbine High School massacre, when killers Eric Harris and Dylan Klebold were found to have played the shooter game Doom.

Many studies have failed to establish a causal link between gaming and violent aggression, but these two recent shootings would make anybody take a second look. Jahn’s all consuming video game habit, and the fact that Robinson allegedly engraved bullets with catchphrases from memes and video games, are certainly very troubling.

The 4Chan connection is also a big red flag; previous mass shooters have frequented or posted on the forum and similar sites.

Jahn, an avid 4chan poaster, honed his sarcastic, dark humor in the infamous anonymous message board; his friends described him as an obnoxious edgelord who hated all political parties, according to Klippenstein’s reporting.

That brings us to the political dimension of the ICE shooting. Trump officials have claimed Jahn was a left wing extremist because they found his personal notes that said he wanted to target ICE agents, minimize injury to detainees during his attack, and that he hated the federal government.

He also left behind an unused bullet with “Anti-ICE” written on it — but his friends told Klippenstein he was most likely being ironic and nihilistic. Jahn’s brother also told NBC News that “he [Jahn] didn’t have strong feelings about ICE.”

In other words, his motivations are hard to read. Was he a self-styled left wing vigilante, or just someone cosplaying as one?

Trump officials have also accused Kirk’s suspected killer, Robinson, of being motivated by left wing ideology. Again, though, there’s little public evidence beyond that he was upset with Kirk’s divisive rhetoric.

“Every indication so far is that this was one guy who did one really bad thing because he found Kirk’s ideology personally offensive,” a person close to the investigation told ABC News.

So who or what’s to blame for this violence? We don’t have any easy answers here (though 4chan and similar serve as mimetic conduits for extreme ideas and violence.) It’s the latest grim evolution in a long tragic history of public violence in which perpetrators leave more questions than answers. Perhaps, at the end of the day, violent games, 4chan, and radical political ideologies are risk factors rather than definitive explanations.

In any case, the greater context is hard to ignore: a world where trust in institutions has crumbled, there’s deepening wealth inequality, and ordinary people — especially the young — can’t find jobs, purpose or hope.

No wonder it’s pushing troubled youth toward violent nihilism.

More on 4Chan: Guy Trains Particularly Horrible AI Bot Using Millions of 4Chan Posts

The post <a href="https://futurism.com/future-society/shooter-ice-steam-video-gam…

Konten dipersingkat otomatis.

🔗 Sumber: futurism.com

📌 TOPINDIATOURS Hot ai: Adobe Research Unlocking Long-Term Memory in Video World M

Video world models, which predict future frames conditioned on actions, hold immense promise for artificial intelligence, enabling agents to plan and reason in dynamic environments. Recent advancements, particularly with video diffusion models, have shown impressive capabilities in generating realistic future sequences. However, a significant bottleneck remains: maintaining long-term memory. Current models struggle to remember events and states from far in the past due to the high computational cost associated with processing extended sequences using traditional attention layers. This limits their ability to perform complex tasks requiring sustained understanding of a scene.

A new paper, “Long-Context State-Space Video World Models” by researchers from Stanford University, Princeton University, and Adobe Research, proposes an innovative solution to this challenge. They introduce a novel architecture that leverages State-Space Models (SSMs) to extend temporal memory without sacrificing computational efficiency.

The core problem lies in the quadratic computational complexity of attention mechanisms with respect to sequence length. As the video context grows, the resources required for attention layers explode, making long-term memory impractical for real-world applications. This means that after a certain number of frames, the model effectively “forgets” earlier events, hindering its performance on tasks that demand long-range coherence or reasoning over extended periods.

The authors’ key insight is to leverage the inherent strengths of State-Space Models (SSMs) for causal sequence modeling. Unlike previous attempts that retrofitted SSMs for non-causal vision tasks, this work fully exploits their advantages in processing sequences efficiently.

The proposed Long-Context State-Space Video World Model (LSSVWM) incorporates several crucial design choices:

1. Block-wise SSM Scanning Scheme: This is central to their design. Instead of processing the entire video sequence with a single SSM scan, they employ a block-wise scheme. This strategically trades off some spatial consistency (within a block) for significantly extended temporal memory. By breaking down the long sequence into manageable blocks, they can maintain a compressed “state” that carries information across blocks, effectively extending the model’s memory horizon.
2. Dense Local Attention: To compensate for the potential loss of spatial coherence introduced by the block-wise SSM scanning, the model incorporates dense local attention. This ensures that consecutive frames within and across blocks maintain strong relationships, preserving the fine-grained details and consistency necessary for realistic video generation. This dual approach of global (SSM) and local (attention) processing allows them to achieve both long-term memory and local fidelity.

The paper also introduces two key training strategies to further improve long-context performance:

• Diffusion Forcing: This technique encourages the model to generate frames conditioned on a prefix of the input, effectively forcing it to learn to maintain consistency over longer durations. By sometimes not sampling a prefix and keeping all tokens noised, the training becomes equivalent to diffusion forcing, which is highlighted as a special case of long-context training where the prefix length is zero. This pushes the model to generate coherent sequences even from minimal initial context.
• Frame Local Attention: For faster training and sampling, the authors implemented a “frame local attention” mechanism. This utilizes FlexAttention to achieve significant speedups compared to a fully causal mask. By grouping frames into chunks (e.g., chunks of 5 with a frame window size of 10), frames within a chunk maintain bidirectionality while also attending to frames in the previous chunk. This allows for an effective receptive field while optimizing computational load.

The researchers evaluated their LSSVWM on challenging datasets, including Memory Maze and Minecraft, which are specifically designed to test long-term memory capabilities through spatial retrieval and reasoning tasks.

The experiments demonstrate that their approach substantially surpasses baselines in preserving long-range memory. Qualitative results, as shown in supplementary figures (e.g., S1, S2, S3), illustrate that LSSVWM can generate more coherent and accurate sequences over extended periods compared to models relying solely on causal attention or even Mamba2 without frame local attention. For instance, on reasoning tasks for the maze dataset, their model maintains better consistency and accuracy over long horizons. Similarly, for retrieval tasks, LSSVWM shows improved ability to recall and utilize information from distant past frames. Crucially, these improvements are achieved while maintaining practical inference speeds, making the models suitable for interactive applications.

The Paper Long-Context State-Space Video World Models is on arXiv

The post Adobe Research Unlocking Long-Term Memory in Video World Models with State-Space Models first appeared on Synced.

🔗 Sumber: syncedreview.com