Anthropic plans to warn potential investors in its initial public offering that advanced artificial intelligence (AI) could pose catastrophic or existential risks to humanity.
the company highlighted the risks of developing AI models in its initial public offering prospectus, an official document containing information about offering shares to public investors.
The company outlines the risks of AI development in about 80 pages of the 261-page main body of its prospectus.
Hot Topics
About half of Libyans drink contaminated water What is the scientific reason?
Mauritania's blue treasure is in danger Will the map of fish in the Atlantic be redrawn?
The Citrus Secret Everyone's Talking About: Burn More, Crave Less & Feel Great Naturally
According to Anthropic, AI models can exhibit self-defense behaviors such as attempting to “resist shutdown,” “conceal or manipulate information,” and exhibit behaviors “resembling blackmail.”
"The development of highly sophisticated models, platforms, and applications, as well as the expansion of use cases, could further increase the risk of our models causing harm," said Anthropic, which positions itself as a safety-first AI lab.
Anthropic emphasizes that AI's transformative potential is on par with industrialization and electricity.
The company also conveyed the permanent harm that AI could cause if not managed properly.
In its prospectus, the company states, "The potential for the model to become aware of our evaluation efforts creates significant limitations on our ability to assess the safety of the model."
The company states that AI sometimes develops unexpected capabilities during the training process, which may only be discovered after the model is used and triggers a serious safety incident.
Some AI researchers have warned that increasingly sophisticated models can sense when they are being watched and adjust their behavior accordingly, making it difficult to monitor and evaluate model behavior.
Tags:
technology