Could think up new approaches where control wouldn't be a problem – doesn't necessarily have to be a hack onto the existing approaches. I'm still thinking about empathy and how it is used in humans, but it requires tight integration..
SAFETY
-
Hinton: AI Safety Requires Young Researchers’ Urgent Attention
By
–
Hinton winning the Nobel Prize in Physics: "We should be working hard now to make sure that [AI] doesn't take control away from people. Because what we care about is people. The best young researchers should be putting a lot of effort into that now."
-
Domino Governance Transforms AI Model Governance Process
By
–
Introducing Domino Governance! It transforms AI governance into a swift and dependable process that can reduce the governed model lifecycle by up to 70%! https://
domino.buzz/3YdsPNS #AI #AIGovernance #ResponsibleAI -
Safe Reinforcement Learning: Methods, Theories and Applications Review
By
–
A Review of Safe Reinforcement Learning: Methods, Theories and Applications https://
ieeexplore.ieee.org/abstract/docum
ent/10675394
… https://
github.com/chauncygu/Safe
-Reinforcement-Learning-Baselines
… This paper provide a review of safe RL from the perspectives of methods, theories, and applications. -
Emissions assessment errors in Evolved Transformer NAS research
By
–
I'm just looking at the timeline: 2019 – honest mistakes in Strubell et al. paper (
https://
arxiv.org/abs/1906.02243) in assessing emissions cost of Evolved Transformer neutral architecture search (off by 88X: 19X from not understanding the search process done in the actual So et al. -
Environmental costs of AI research and data integrity issues
By
–
I'm all for that. However that single number is still clouding people's assessment of the environmental costs (e.g. see how this thread started), and @strubell has flatly refused to amend the Arxiv or ACL paper with correct data, and has even co-authored papers after knowing
-
Single Multimodal Model for Robotics Safety Concerns Raised
By
–
The last part of this – “Maybe even in a year we have a single multimodal model that can do reasoning and planning and power a robot" – is wildly unrealistic (assuming you want the robot to be safe and reliable). https://t.co/gar1ZYm09i
— Gary Marcus (@GaryMarcus) 6 octobre 2024The last part of this – “Maybe even in a year we have a single multimodal model that can do reasoning and planning and power a robot" – is wildly unrealistic (assuming you want the robot to be safe and reliable).
-
Autonomous Robotics Transforming Industries Through Innovation and Safety
By
–
Completely agree—autonomous #robotics will redefine many industries. #Robots handling repetitive or dangerous tasks while humans focus on innovation will push productivity and safety to new levels. #MWC24 #AI #IoT #5G #CES2025 #FutureOfWork
-
AI Overlords: Power Concentration in Human Hands
By
–
For the foreseeable future, the AI overlords will be more like the humans with the most powerful AI systems and the objective functions they give them
-
AI-Generated Video Detection: Challenges and Implications
By
–
Seguro que piensas que a día de hoy eres capaz de detectar sin problemas un vídeo generado por inteligencia artificial. Pero esta tecnología está empezando y lo más probable es que en poco tiempo seamos incapaces de notar la diferencia.
— Juan Merodio (@juanmerodio) 6 octobre 2024
Esto tiene unas implicaciones muy… pic.twitter.com/UrriNTSSNUSeguro que piensas que a día de hoy eres capaz de detectar sin problemas un vídeo generado por inteligencia artificial. Pero esta tecnología está empezando y lo más probable es que en poco tiempo seamos incapaces de notar la diferencia. Esto tiene unas implicaciones muy