I'm arguing with someone on the Internet and would like you to review my next post…
- User
I'm arguing with someone on the Internet and would like you to review my next post before I hit send: "Despite having invented the transformer architecture in 2017, Google did literally nothing with it until OpenAI made ChatGPT public in November 2022, forcing Google's hand. For the want of ad revenue, Google held back AIs from the public for about five years."
Assistant
I'll help you review this statement. While there are some elements of truth, the claim is somewhat oversimplified and not entirely accurate:
-
Google was actually actively working on transformer and AI technologies during this period. They published significant research and developed models like BERT (2018), T5 (2019), and LaMDA (2021).
-
Google did release some AI products and research during this time, including Google Assistant improvements, search enhancements, and research publications.
-
The statement overlooks Google's substantial contributions to the transformer architecture and AI development.
A more nuanced statement might acknowledge Google's ongoing AI work while recognizing that they were slower to release a conversational AI product compared to OpenAI.
If you want to make a strong point, you could focus on the specific ways Google was cautious about public AI releases, rather than claiming they "did literally nothing" or suggesting it was purely about ad revenue.
Would you like help crafting a more precise and factually supported statement?
-
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.