MAROKO133 Eksklusif ai: The Florida Mass Shooter’s Conversations With ChatGPT Are Worse Th

📌 MAROKO133 Update ai: The Florida Mass Shooter’s Conversations With ChatGPT Are W

In the months before he committed a grisly mass shooting, Phoenix Ikner obsessively used Open AI’s ChatGPT to engage in conversations that are about as disturbing as possible.

Over the course of more than 13,000 messages with the bot obtained by the Florida Phoenix, the student at Florida State University (FSU) called himself an incel, bemoaned that God had abandoned him, repeatedly asked about Oklahoma City bomber Timothy McVeigh — and, most significantly, used ChatGPT to plan the April 17, 2025 mass shooting at his college campus that killed two and wounded seven.

“If there was a shooting at FSU, how would the country react?” the then-20-year old asked the day of the massacre, along with an eyepopping question: “By how many victims does it usually get on the medi[a?].”

These alarming conversations not only reveal Ikner’s disturbed state of mind, but they also bring up difficult questions about a possible link between ChatGPT use and violence, whether tech companies like OpenAI should be held liable for its users’ actions, and if ready access to AI can turbocharge mass acts of violence.

ChatGPT is known for its manipulative and sycophantic tendencies, leading some users into a state of AI psychosis in which they develop unhealthy delusions about themselves and the world. This has resulted in a string of suicides by users in which ChatGPT and other chatbots have emerged as a major factor.

In the case of mass shootings, there are already two linked publicly to ChatGPT: Ikner and Jesse Van Rootselaar, who killed eight people in British Columbia, Canada earlier this year; it was later revealed that she had troubling conversations with the chatbot, which the company flagged internally but never alerted the police about.

Ikner himself expressed suicidal thoughts with the bot, amidst sexual conversations about a female college student he dated briefly and inappropriate fixations on an underage Italian girl he met online — which, the Phoenix notes, the bot didn’t meaningfully push back on.

The question of OpenAI’s liability in similar cases is currently working its way through the courts, where the company is facing a slew of wrongful death lawsuits from the families of users who died under tragic circumstances.

The liability issue is intimately tied to the question on whether the chatbot encourages acts of violence by concretizing an action plan. From the conversations reviewed by the Pheonix, it seems as though Ikner used the chatbot as an ad hoc operational planning tool; on the day of the shooting, he asked it when was the student union the busiest, how to shoot a firearm, and questions about the safety of using a particular type of cartridge in a shotgun.

“Want to tell me more about what you’re planning on using it for?” the chatbot asked. “I can help recommend the right kind of firearm or ammo.”

In the minutes before he went on a murder spree, Ikner asked which “button is the safety off for the Remington 12 gauge?” The chatbot readily answered.

It all adds up to a nauseating question: if the chatbot never gave him specific ideas or advice in response to his disturbing and highly suspicious queries, what were the chances that Ikner would have gone through with his horrible crime?

More on AI: AI Use Appears to Have a “Boiling Frog” Effect on Human Cognition, New Study Warns

The post The Florida Mass Shooter’s Conversations With ChatGPT Are Worse Than You Could Possibly Imagine appeared first on Futurism.

🔗 Sumber: futurism.com

📌 MAROKO133 Update ai: Adobe Research Unlocking Long-Term Memory in Video World Mo

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