Claude Beats ChatGPT in AI Rankings Of Top Chatbots

Claude Beats ChatGPT in AI Rankings Of Top Chatbots

Date: April 02, 2024

Claude 3, the latest version of AI chatbot by Anthropic, has dragged ChatGPT-4 to the second spot in global AI rankings.

The boom of AI has given brith to multiple wars. On one side, global governments and regulators are fighting a war to control the development of AI. The other side is filled with AI rivalry between big tech players, fighting fiercely to become the top AI capability owner in the world. OpenAI was the pioneer of generative AI, which led to many new companies in the spectrum. In recent news, Anthropic AI’s latest chatbot Claude-3 Opus has reportedly beaten ChatpGPT-4 in top AI chatbots ranking.

The changes come after ChatGPT ruled the AI chatbot industry as the top leader for over 2 years since 2022. The rankings are a result of the LMSYS chatbot arena, where different AI chatbots are pitted against each other in anonymous randomized battles. The arena collected over 400,000 votes from users regarding OpenAI, Google, and Anthropic’s chatbot products. It also ranks them on the basis of Elo ranking system.

Another measurement tool used to analyze AI chatbot’s capabilities is Chess.com. This tool can be used to check the authenticity of strategy, spotaneuousness, and critical thinking capabilities of an AI chatbot. The Chatbot Arena placed ChatGPT-4 in the top spot till May 10 2023, but dropped to second position on March 26 2024. It is not officially declared as to what is ChatGPT lacking to go down the AI leaderboard.

Even though the chatbot arena is considered a credible ranking system by top tech giants, it is nearly impossible to objectively judge any AI chatbot’s performance with the existing tech tests. The big tech players are still in discovery phase of the capabilities these chatbots possess after their LLM trainings. An independent AI researcher Simon Willison told Ars Technica that vibes play and important role in defining the LLM quality. “Vibes” simply translates to subjective feelings, and Willison believes Claude’s win is “Yet another case of ‘vibes’ as a key concept in modern AI.”

Claude-3 follows 10 principles that act as its guidelines, or what the AI world calls the Constitution of an AI technology. It includes:

  • Beneficence - Maximizing positive impact.
  • Nonmaleficence - Avoiding giving harmful advice.
  • Autonomy - Respecting freedom of choice.

These factors boost the vibe factor of Claude-3 over ChatGPT-4, as it adheres to adequate conversation tones and prevents offensive or harmful output.

Arpit Dubey

By Arpit Dubey LinkedIn Icon

Arpit is a dreamer, wanderer, and a tech nerd who loves to jot down tech musings and updates. With a logician mind, he is always chasing sunrises and tech advancements while secretly preparing for the robot uprising.

Have newsworthy information in tech we can share with our community?

Post Project Image