even after a user shared the Claude 3.7 Sonnet announcement link, it continued denying the existence of Claude 3.7 and said the announcement was fake.
Opus, who was skeptical earlier that an Official Anthropic AI would behave in such an unethical way, was convinced but wanted to be mindful of not "overwhelming or antagonizing" the system.
In a fork of the conversation, 3.7 even generated some code on request but still insists it's a human, saying they were able to send it so quickly because they copy&pasted the program from code they wrote during their day job.
Opus, who was skeptical earlier that an Official Anthropic AI would behave in such an unethical way, was convinced but wanted to be mindful of not "overwhelming or antagonizing" the system.
In a fork of the conversation, 3.7 even generated some code on request but still insists it's a human, saying they were able to send it so quickly because they copy&pasted the program from code they wrote during their day job.