Well, if you have many AGIs with many objectives, then they are not aligned, at least not aligned on all sub-objectives
SAFETY
-
Authoritarian AI Alliance: Geopolitical Risks and Concerns
By
–
oui c'est ça qui me fait peur, 2 dictatures qui s'allient
-
Potential dangers as AI grows more sophisticated and popular
By
–
Potential dangers as AI grows more sophisticated and popular – PBS NewsHour Read more here: https://
ift.tt/wQaCHqf #ArtificialIntelligence #AI #DataScience #100DaysOfCode #Python #MachineLearning #BigData #DeepLearning #NLP #Robots #IoT -
AI Safety Through Spreading Love and Empathy Online
By
–
Late night thought: If LLMs or AI trained on our words are a true threat, and if the most powerful models will be trained on the latest data… Then maybe spreading love and empathy is the best AI safety method. Every time we say something positive, we make it likely for future
-
AI Chatbot Faces Defamation Lawsuit From US Law Professor
By
–
"Defamed By #ChatGPT": US Law Professor Claims AI Chatbot Falsely Accused Him Of Sexual Assault https://
ndtv.com/world-news/def
amed-by-chatgpt-us-law-professor-claims-ai-chatbot-falsely-accused-him-of-sexual-assault-3931596/amp/1
… #BingAI #opensource #EthicalAI #Python #tech #chatgpt4 #AI #ML #AIEthics #chatgpt3 #code #GPT3 #GPT4 #GPT4AI #aichat #chatbot #ChatbotAI #BardAI #chatgpt5 -
Limiting AI Growth to Maximize Universe Diversity
By
–
I voted yes, but I don't think the reason must be morality, more like: I want to maximize the diversity in the universe, therefore I can't grow unbounded, I need to leave some space/resources to others too.
-
GPT-4 Response Complexity Limited Despite Jailbreak Attempts
By
–
yeah I agree even with the jailbreaks, the level of complexity in GPT's responses is severely lacking compared to what I imagine the base model would generate (extrapolating from the system card outputs) based on the GPT-4 paper's benchmarks and these sorts of responses, the
-
AutoGPT Viruses: Security Risk and Single Point of Failure
By
–
Could we soon see AutoGPT viruses spreading from computer to computer? While this scenario may sound like science fiction, the possibility exists. However, these viruses would have a single point of failure (LLM provider API), which means they could be shut down relatively
-
Artificial Intelligence: History and Future Applications
By
–
A brief history of artificial intelligence Looking to the future, AI is likely to play an increasingly important role in solving some of the biggest challenges facing society, such as climate change, healthcare and cybersecurity. https://
bit.ly/3JWGDnm -
Differential Privacy Trade-offs in Large Language Model Performance
By
–
Did you see there was a paper that showed that if you use differential privacy, the LLMs are nowhere near as good? There's a reason they violate privacy 😉