GPT-3-175B is almost 3 years old. Today a 3B model – 60x smaller – can achieve comparable performance. What happens if today's cutting edge models are on that trajectory? …they will proliferate far and wide… this is the containment problem in a nutshell
REGULATION
-
Japan Addresses AI Property Rights and Intellectual Property Concerns
By
–
This is interesting, as I just chatted with my friends a few days ago that in Japan the AI related property issues are seriously concerned..
-
Distinguishing Urgent AI Safety Guardrails from Long-term Existential Regulation
By
–
People often conflate what’s important vs. urgent, like right now, every lawmaker in the world thinks you need to regulate to avoid existential risk from #AI, it’s important but the urgent task is to have guardrails to ensure safe deployments of #AI!
-
Lawmakers Introduce Bill to Regulate AI Safety and Nuclear Risk
By
–
Lawmakers Introduce Bill to Keep AI from Going Nuclear
#AI #AIio #BigData #ML #NLU #Futureofwork @CRudinschi @AntonioSelas @alexjc @RobotLaunch @karpathy @andyjankowski @bobgourley @CadeMetz http://
ow.ly/SafV30svtqk -
Microsoft Developing Private ChatGPT for Dedicated Servers
By
–
For businesses concerned about #dataleaks and #regulatorycompliance, #Microsoft may be building a private #ChatGPT for dedicated servers. The cost could be 10x more than the regular offering. (#OpenAI may be doing this too.) #GenAI #AI #privacy
-
AI Fear-Driven Regulation Risks Net Negative Outcomes
By
–
I’m less worried that difficulties lead to paralysis. I’m more worried that this death by AI fears will lead to frantic regulation that is a net negative.
-
US regulatory study on AI and ChatGPT rules begins
By
–
US begins study of possible rules to regulate #AI like #ChatGPT https://
reut.rs/44GpdVW #ethics -
AI-Controlled Drone Kills Operator in USAF Simulated Test
By
–
“We trained the system–‘Hey don’t kill the operator–that’s bad. You’re gonna lose points..’ So what does it start doing? It starts destroying the communication tower that the operator uses to communicate with the drone to stop it from killing the target.” https://
vice.com/en/article/4a3
3gj/ai-controlled-drone-goes-rogue-kills-human-operator-in-usaf-simulated-test
… -
Data Availability Statements in Scientific Journals: The Accessibility Challenge
By
–
Data is not available upon request https://
bit.ly/3MJ472b
Many journals now require data sharing and require articles to include a Data Availability Statement. However, … #AI #MachineLearning #DeepLearning #LLMs #DataScience