Release of Advanced AI Tool to Public
A version of an artificial intelligence (AI) tool, which the company previously stated was too powerful for public release, has now been made available to the public.
Claude Fable 5 is a variant of Anthropic's Claude Mythos, an AI program that attracted significant attention among leaders in technology, finance, and government when it was initially released privately in April for preview and testing purposes.
Concerns were raised by some senior figures about the tool's potential to pose financial security risks due to its advanced capabilities, although others have questioned whether the concerns were amplified by marketing narratives.
On Tuesday, Anthropic announced that Fable would be released with safeguards and user limitations, acknowledging that
"releasing a model this capable comes with risks". The company further stated,
"Fable's capabilities exceed those of any model we've ever made generally available".
Initial Private Release and Company Valuation
When Mythos was first made available to a limited number of organizations for preview, Anthropic explained that the tool's intelligence level was such that it could be potentially dangerous.
Anthropic is anticipated to become a publicly traded company soon, with its private valuation approaching $1 trillion (£747 billion).
Expanded Access to Claude Mythos 5
Additionally, Anthropic revealed on Tuesday that approximately 150 groups previously granted access to preview Mythos will now have access to Claude Mythos 5. This version does not impose limitations on cybersecurity or biology, depending on the specific applications of the organizations.
The company clarified that this access is currently limited to a
"small group of cyberdefenders and infrastructure providers", but it expects to broaden this access in the near future.
"We intend to expand access to Mythos 5 through a broader trusted access program", Anthropic stated.
Capabilities and Operational Features
Both Fable and Mythos, which are essentially the same model but differ in safeguards and access levels, can operate
"unattended" on human commands that the tools are given for longer periods of time "than any previous Claude models", Anthropic added.
Industry Perspectives on AI Advancement
Anthropic co-founder Jack Clark recently spoke to night, expressing concerns about the rapid expansion of AI capabilities. He suggested that the public should have mechanisms to slow the pace of technological advancement.
"You want the option to be able to take your foot off the gas and put your foot on the brake", Clark said.
"Right now, it's like the AI industry has a gas pedal, but it doesn't have a brake pedal."
