We are here, my babies. A canon event.
The latest model from Anthropic, dubbed Mythos, has been scaled back and kept from being released to the public. The model has been called “too powerful” and a danger to systems worldwide. Instead, Anthropic created an initiative dubbed “Glasswing” to give major companies and organizations access to the model so they can use it to better patch and protect their environments.
Let’s be real, this is newsworthy for the reasons you wouldn’t expect. Anthropic decided this AI model was too powerful to be unleashed on the general public and held it back, choosing to donate 100 million dollarydoos worth of credits and 4 million to nonprofit organizations. I believe OpenAI is working on their own very powerful model. Will they hold it back too, or release it?
Why is it too powerful, I ask. Anthropic answers that it found many zero-day bugs, some going back 27 years. If this had been found by threat actors, it would be a global security event. No system would be able to defend itself when all those attacks are unleashed.
As high as the thigh highs go, so does my belief that one day these systems will be so powerful that no amount of patching, security, or configuration would be able to stop an advanced decryption machine. An apocalyptic event for sure, so hopefully I’m wrong.
