Post

Finally, an agentic AI that reliably finishes a multi-step task instead of getting stuck in a loop and suggesting you eat rocks.

Are AI executives begging for immunity from damage they created with hallucinations? Is the sudden concern about "humanity" a smokescreen for a technology with a 95% failure rate? Have AI executives been making knowingly false promises for years (i.e. fraud)? www.thedataexperts.us/writing/ente...

Interesting that agentic tools are now part of how attackers work. Which kinds of guardrails do you think are the hardest to maintain at scale?

I'd love a companion test on the defender's side: the same budget, the same target, and agents trying to harden the system. The useful comparison is how the balance changes, beyond either demo.

IT security specialists around the world are all-in on using agentic AI to harden their systems -- and relying on tools like Anthropic's Claude to help them be more effective at their jobs. Meanwhile, It's mostly open weight Chinese systems that are used to stage attacks, as Anthropic noted.

is this an actual picture? there is a real artist who has the signature "Zimbo" from what i've seen but they dont draw like this at all

Yes, running head first into a wall chasing an illusion has side effects.