The next wave of the curve has hit in the midst of IPOs filed and familiar players finally joining the fray.
Anthropic's cybersecurity model, Mythos, has a public-facing version: Claude Fable 5. From the frontier lab itself,
"Mythos-class models excel at discovering and exploiting software vulnerabilities. They can thus make cyberattacks substantially easier and cheaper to commit. Mythos-class models also show strong skills in agentic hacking. This involves performing multiple different parts of a cyberattack in addition to finding exploits—reconnaissance, discovery, lateral movement, and more..."
Basic tier users of Claude can access Fable as part of their subscription for the next two weeks.
I wasn't expecting this drop to accelerate things that much.
Combine the over-hyped nature of the Mythos saga with my decreased use of Claude Code as of late, and you end up with a writer who thought we'd reached the exponential's end or, at least, the fruit we'd receive from that exponential's end.
In the last few hours since its release, I've thrown all my go-to queries, standards, and codebases at it to benchmark, get a feel for, and test its abilities on "finding exploits" and improving my work through its agentic coding capabilities.
I'm again blown away.
Leave aside all potential world-ending capabilities for the moment (Anthropic warns that "Distillation of Fable 5’s abilities could indirectly lead to the proliferation of near-frontier AI capabilities—and these could be released without the appropriate safeguards..." Nice.) and turn to its application for the hobbyist, SMB owner, entrepreneur, or artist: genuine savings and increased capabilities to build on digital projects and online infrastructure.
Or, in my case, a back-linked index of all my writing, cybersecurity fixes, and a (massive) 14% reduction in the messy and bloated codebase that was starting to become a genuine problem for reading work on my site.
I continue to wrestle with the implications of using this technology.
I hear the angst and ennui of this generation and feel it myself.
I am deeply concerned about the incentives driving these labs.
As I write through all the creative inspiration and empowerment I've found in these tools though, I can't help but think that the worst sides of the story muddy the waters too much...
Update 06/13/26: Well that was not entirely unexpected...