Anthropic built a democratic AI chatbot by letting users vote for its values

In the Collective Constitutional AI experiment, Anthropic tried to integrate group-based feedback into the models constitution. As this experiment appears to be the first of its kind, and it relies on Anthropics Constitutional AI approach, there isnt an established test for comparing base models to those tuned with crowd-sourced worths. Ultimately, it appears as though the model that implemented data resulting from user polling feedback surpassed the base design “somewhat” in the location of biased outputs.

Thank you for reading this post, don't forget to subscribe!

In what may be a first of its kind study, expert system (AI) company Anthropic has developed a big language design (LLM) thats been fine-tuned for valuation by its user community.What does it imply for AI advancement to be more democratic? To find out, we partnered with @collect_intel to use @usepolis to curate an AI constitution based on the viewpoints of ~ 1000 Americans. We trained a design against it utilizing Constitutional AI. pic.twitter.com/ZKaXw5K9sU— Anthropic (@AnthropicAI) October 17, 2023

In what may be a very first of its kind research study, synthetic intelligence (AI) company Anthropic has established a big language design (LLM) thats been fine-tuned for worth judgments by its user community.What does it mean for AI development to be more democratic? In the Collective Constitutional AI experiment, Anthropic attempted to integrate group-based feedback into the designs constitution. As this experiment appears to be the very first of its kind, and it relies on Anthropics Constitutional AI approach, there isnt a recognized test for comparing base designs to those tuned with crowd-sourced values. Ultimately, it appears as though the model that executed data resulting from user polling feedback outshined the base design “slightly” in the location of prejudiced outputs.