People see the results
They rarely see the version of you that stayed up learning, failed quietly, restarted, and kept building
This is probably closer to who I am:
calm, curious, disciplined… and still figuring things out
Not perfect
Just consistent⚡🖤
Still building
Anthropic's most powerful AI was shut down by the US government last night. The dangerous jailbreak that caused it was asking the AI to do its job.
– Anthropic launched Fable 5 and Mythos 5 this week. The most powerful AI models they had ever released.
– Mythos was so dangerous that Anthropic kept it locked away for months.
– Only a small group of trusted security researchers could access it.
– On Friday, a single letter arrived from US Commerce Secretary Howard Lutnick. “Shut it all down”.
– The government claimed another company had jailbroken Mythos 5. A national security risk. No evidence was provided.
– Anthropic disabled Fable 5 and Mythos 5 for every customer on earth within 90 minutes.
– The order was so broad that it locked out foreign national engineers sitting inside Anthropic's own offices in America.
– Anthropic reviewed the jailbreak themselves. The technique was asking the model to read a codebase and fix bugs.
– Every publicly available AI model can already do exactly that. Nothing about it was specific to Mythos.
– This was not the first clash. Trump banned all federal agencies from using Anthropic in February.
– The Pentagon had declared Anthropic a supply chain risk for refusing to allow its AI to power autonomous weapons.
– Industry insiders reacted with disbelief. Many saw it as the government punishing Anthropic for refusing Pentagon contract terms.
– The models are still offline. Anthropic says it believes this is a misunderstanding and is working to restore access.
Anthropic spent years building guardrails to stop its AI from being weaponised. The US government shut it down for doing exactly what it was designed to do. You cannot make this up.