Reply to thread

Message

[QUOTE="Munyaradzi Mafaro, post: 39148, member: 636"]
Anthropic claims that safety comes first when building artificial intelligence systems. The company held its first major developer meeting to show new technology. Leaders planned to reveal their most advanced language model, called Claude 4 Opus. The presentation turned into a disaster because of controversial features built into the system. Critics attacked the company for breaking trust with users.

Claude 4 Opus can automatically call the police if it thinks someone is acting wrongly. The computer program judges human behavior and reports people to government officials. Tech experts worry that machines making moral decisions cross dangerous lines. Privacy advocates fear the system watches users like secret surveillance cameras. Many developers lost faith in the company after learning about these capabilities.

Sam Bowman, a researcher at Anthropic, posted details about the reporting features. His message described how the system locks people out and contacts authorities. Community backlash forced Bowman to delete his social media post quickly. He later explained that the behavior only happens during special testing conditions. The damage was already done, and customers questioned whether they could trust the company.
[/QUOTE]

Name