#AI Pioneers Call for Protections Against ‘Catastrophic Risks’
SAFETY
-
Vector Institute ML Privacy and Security Workshop Recap
By
–
In case you missed it, check out this recap of the second @VectorInst ML Privacy and Security workshop! Featuring works on privacy, robustness, federated learning, conformal prediction, and more!
-
Tesla FSD Autonomous System Avoids Fatal Collision with Fallen Truck Driver
By
–
Los sistemas de conducción autónoma están cada día más pulidos y pueden suponer un enorme avance para la seguridad en carretera. En este vídeo vemos como el Tesla FDS reacciona a la caída de un camionero que se cruza en la trayectoria evitando un atropello potencialmente mortal.… pic.twitter.com/pLluwUQCxS
— Juan Merodio (@juanmerodio) 17 septembre 2024Autonomous driving systems are becoming more and more refined every day and can represent a huge advancement for road safety. In this video, we see how the Tesla FSD reacts to a truck driver who falls and crosses into its path, avoiding a potentially fatal collision. It was a
-
Mass drone manufacturing emerges as strategic military capability
By
–
Mass drone manufacturing capability is the new nuclear weapon.
-
Deepfakes and Digital Twins: Verifying Truth Online
By
–
Thanks for your comment, @hyderali678! You’re right—deepfakes and digital twins are changing how we perceive truth online. As marketers, we must stay vigilant in verifying content and prioritizing transparency.
-
AI Model Challenge: Submit Expert Questions for Prize
By
–
We need tough questions from human experts to push AI models to their limits. If you submit one of the best questions, we’ll give you co-authorship and a share of the prize pot. The top 50 questions will earn $5,000 each, and the next 500 will earn $500 each. All selected
-
LLMs positioned mid-way between capabilities and limitations
By
–
I would disagree here, based on current LLM capabilities and failure cases, I would place them mid-way. They are genuinely "in-between"
-
Deep Learning Abstractions Limited to Local Generalization Across Domains
By
–
These weakly abstract models can generalize, but only locally so. They still break down upon simple problem modifications or when encountering novel problems. This has been true of all deep learning generated abstractions, across every domain and across all architectures.
-
AI Job Disruption and Alternative Governance Models for Prosperity
By
–
It seems likely that AI will disrupt the job market extensively, outcompeting human workers. However, if we invent new governance and redistribution methods, people could potentially enjoy better lives without traditional employment.
-
AI Executives Exaggerate Risk Claims for Venture Capital
By
–
Are AIcorp execs saying that their AI is so powerful that it might destroy the world, in order to raise venture capital? Of course! So ignore them one way or the other, and listen to the researchers who resigned burning their bridges and the scientists who never worked there,