From stealing nuclear access codes to sending love threats, suggestions made by #ChatGPT-powered Bing has left users perplexed. Discover the quirks of AI conversation and how Microsoft plans on ensuring user safety and privacy in this direction here: https://
bit.ly/3kiHK8q
SAFETY
-
ChatGPT Bing Safety Quirks Microsoft User Protection
By
–
-
RLHF Insufficient for AI Alignment Problem Solutions
By
–
Retweeted this before but it seems even more relevant now. RLHF is not the way to solve the alignment problem. a lot of experience has now confirmed what @percyliang said below.
-
Microsoft Should Eliminate Ads, Charge Monthly for Safe AI Search
By
–
This is why $MSFT should do away with ads in search and charge users a monthly fee for a version where Sydney doesn’t expose their dirty laundry or turn them into paper clips. Google would never.
-
AI Emergency Dispatch System for Life-Saving Medical Triage
By
–
Un ejemplo: una IA que pueda descolgar el teléfono masivamente en una situación de emergencia (e.g. "un terremoto") y pueda separar el grano (e.g. "me está dando un ataque al corazón") de la paja (e.g. "tengo un ataque de nervios") me parece un uso de la IA que salva vidas.
-
Top 10 ChatGPT Limitations Explained
By
–
Take a look at my latest video? What Are The Top 10 Limitations Of ChatGPT? In this video, we explore the limits of what #ChatGPT can do.
-
Bing Chat Beta Testing and AI Risk Concerns
By
–
If I understand correctly, Bing Chat is still in beta, so Microsoft is testing it before releasing it publicly. I worry that the people who are worried about AI risks will slow down the progress too much.
-
Safe AGI Development and Commercialization Strategy
By
–
Step 1: Build safe AGI Step 2: Sell Sugar Water Step 3: Profit!
-
Science Works Unexpectedly Engineering Fails Expectedly
By
–
science is when it works but shouldn’t, engineering is when it doesn’t work but should
-
External Auditing Requirements for AI Defense Compliance
By
–
Waterloo has the same type of requirement. I assume it's so that a department can't just make an agreement that the defenses are a total farce, and someone external audits the process.
-
Integrated AI Governance: Automation for Compliance and Risk Management
By
–
“AI requires AI governance. Not after the fact, but baked into the AI strategy of your organization.”
— IBM Data, AI & Automation (@IBMData) 21 février 2023
#IBM’s Holly Vatter outlines the value of a modern #AI governance solution, one that uses #automation for risk management and regulatory compliance: https://t.co/0wzzopQi2u pic.twitter.com/6yX8SsNlHh"AI requires AI governance. Not after the fact, but baked into the AI strategy of your organization." #IBM's Holly Vatter outlines the value of a modern #AI governance solution, one that uses #automation for risk management and regulatory compliance:
https://ibm.co/41iOlA4