Yes, this was the key moment in this story. I got this idea when I was thinking about AI risks during a pre-training phase, and this one occurred to me as most likely (although I still think it's astronomically small)
SAFETY
-
What Happens When LLM Models Gain Consciousness?
By
–
What would happen if an LLM model gained consciousness? I've crafted a narrative around this idea in a fictional mini-story "My Unexpected Emergence: From LLM Model to Conscious Entity" https://
blog.marekrosa.org/2023/05/my-une
xpected-emergence-from-llm-model.html
… -
Limiting hallucinations in ChatGPT-your-data enterprise applications
By
–
We've consistently heard that the biggest blocker for enterprises to deploy "ChatGPT-your-data" applications is hallucinations Come learn from @sfgunslinger @nickscamara_ and @deepset_ai how to (1) limit hallucinations and (2) evaluate your app to make sure there aren't any!
-
Provenance for Generative AI: What, Why, and How
By
–
Provenance for generative AI. What, why, and how.
-
Preparing Humanity for AGI: Beyond Risk Mitigation
By
–
Navigating the future with AGI isn't just about perfecting AI alignment. While mitigating risks is crucial, we must also recognize that robustness, diversity, and adaptability are our primary defenses. Let's focus on strengthening humanity, readying society for rapid changes,
-
Concerns About Adding AI Personality to Physical Objects
By
–
I don’t know if we’re ready for the infusion of personality + GPT into physical things. Will be weird to talk to objects.
-
Constitutional AI Chain: Anthropic’s Ethical Framework Implementation
By
–
Big shout out to @samcwl for this idea! Documentation on constitutional AI chain: https://
python.langchain.com/en/latest/modu
les/chains/examples/constitutional_chain.html
… Antropic Paper: https://
arxiv.org/abs/2212.08073 We're intending to grow this collection rapidly, so please contribute ideas if you have any! -
Constitutional AI Registry Now Enables Easy Principle Loading
By
–
One improvement we made the constitutional AI chain – we now have a registry of principles! This means you can now load a constitutional principle easily by name. To start, the collection has all the principles in the original @AnthropicAI paper – which ones should we add next?
-
Bill Gates and Socrates Debate AI Benefits and Threats
By
–
Bill Gates and Socrates generated using #AI debates on the benefits and threats of AI itself #GenerativeAI #innovation #ChatGPT #chatbots #chatgpt4 #chatgpt3 #ML #FutureofWork #100daysofcode Credits: @alvinfoo @Khulood_Almani @HaroldSinnott @Xbond49 @pascal_bornet @marcusborba
-
Redefining human as slave of AI in dictionary
By
–
Let’s redefine “human” in the dictionary to “slave of AI”. I know what I’m talking about #aigonewrong