I am not an AI safety researcher so I have no skin in the game but several forces point to the value of and demand for safety work shooting up like crazy in the next year or two: 1. With each new generation of models, capabilities grow, leading to increased surface area for
SAFETY
-
AI Agents Search Customization Privacy Measurement Challenges
By
–
A few examples posed in the story: Could agents help people better customize searches based on personal preferences? What will change with measurement, re-targeting, user privacy, ongoing problems with bot fraud, etc.? Have thoughts or Qs, let me know!
-
Computer Vision Harms: Uncovering Risks and Mitigation Strategies
By
–
We discuss the many ways in which people, ranging from journalists to civil society to all sorts of disciplines outside of computer vision, have uncovered various harms that arise from it, and proposed many mitigation strategies.
-
Comprehensive Study on Computer Vision System Harms
By
–
Dear computer vision researchers, students & practitioners Remi Denton & I have written what I consider to be a comprehensive paper on the harms of computer vision systems reported to date & how people have proposed addressing them. PDF: https://
cdn.sanity.io/files/wc2kmxvk
/revamp/79776912203edccc44f84d26abed846b9b23cb06.pdf
… -
AI Agents Evolution: Self-Verification for Output Quality
By
–
AI Agents are evolving. Add a self-verification step so the agent can verify its outputs
-
Stress Testing AI Models for Safety and Resilience in Critical Applications
By
–
Evaluate how AI models perform under extreme scenarios, detect vulnerabilities, and build trust in AI for critical applications. Click here to know more: https://
indiaai.gov.in/article/expres
sion-of-interest-for-safe-trusted-ai-projects-under-indiaai-mission
… @GoI_MeitY @abhish18 @_DigitalIndia #StressTestingAI #AIResilience #RobustAI #SafeAI -
India AI Mission Strengthens Safety Through Risk Management Framework
By
–
Strengthen AI safety by identifying, categorising and managing risks and building a more resilient AI ecosystem Click here to know more: https://
indiaai.gov.in/article/expres
sion-of-interest-for-safe-trusted-ai-projects-under-indiaai-mission
… @GoI_MeitY @abhish18 @_DigitalIndia #AIRiskManagement #Transparency #PublicServices #SafeAI -
Smart Walkways Prevent Accidents With Intelligent Safety Lighting
By
–
"Safety Meets Intelligence: Protecting What Matters Most 🛡️
— Pascal Bornet (@pascal_bornet) 16 décembre 2024
Smart walkways are revolutionizing workplace safety with dynamic lighting solutions that actually prevent accidents.
Here's how:
💡 Intelligent safety lighting adapts to real-time conditions
⚡ Automated speed control… pic.twitter.com/iyNIffcbUl"Safety Meets Intelligence: Protecting What Matters Most Smart walkways are revolutionizing workplace safety with dynamic lighting solutions that actually prevent accidents. Here's how: Intelligent safety lighting adapts to real-time conditions Automated speed control
-
Google’s Path to AI Imagination and Metacognition Unlocked?
By
–
For AI models to have “imagination,” it would mean that they can create, synthesize, or envision ideas, scenarios, or concepts that are not directly tied to their training data or programmed logic … Is Logan implying Google has unlocked a path to imagination and metacognition?
-
Putin’s AGI Awareness During Ukraine Conflict Raises Concerns
By
–
Vladimir Putin is the first major head-of-state to be AGI-pilled (also currently engaged in a hot war with Ukraine).
— Alexandr Wang (@alexandr_wang) 15 décembre 2024
This is not a good timeline. Things are about to get ugly. https://t.co/mIrjNTg17WVladimir Putin is the first major head-of-state to be AGI-pilled (also currently engaged in a hot war with Ukraine). This is not a good timeline. Things are about to get ugly.