The company, founded by several former OpenAI employees, promoted its language model as «better than ChatGPT», but there is a nuance.
The fact is that Claude 3 outperforms only the very first GPT-4 model released by OpenAI a year ago, not the much more powerful GPT-4 Turbo, which in Promptbase’s comparative analysis still beats its competitor in every test.
Somewhat interesting advertising choice from Anthropic, comparing their newly released Claude 3 to GPT-4 on release (March 2023).
According to Promptbase’s benchmarking, GPT-4-turbo scores better than Claude 3 on every benchmark where we can make a direct comparison. https://t.co/k3tQEpTxSc pic.twitter.com/W3sRcOud4f
— Tolga Bilge (@TolgaBilge_) March 4, 2024
Anthropic’s marketers decided instead that it was better not to mention this in the announcement, focusing only on comparative data regarding the «old» GPT-4 (after all, they put this «inconvenient» detail in the announcement notes).
This somewhat inconvenient detail is briefly acknowledged in the footnotes of Anthropic’s announcement:https://t.co/R7PJIcb6Gx pic.twitter.com/Jsxp60qYNk
— Tolga Bilge (@TolgaBilge_) March 4, 2024
In general, it is not the first time that AI developers have resorted to advertising stunts to promote their creations — Google has previously admitted that the Gemini AI demo video was staged.