I agree. The talking dog analogy deflates those claims while still pointing out what is unique and worth following up on about the results. Meanwhile, the cho

I agree. The talking dog analogy deflates those claims while still pointing out what is unique and worth following up on about the results. Meanwhile, the chorus of "look this AI still makes dumb mistakes and is not AGI" takes has gotten louder in many circles than the marketing drumbeat. It risks drowning out actual progress and persuading sensitive researchers to ignore meaningful ML results, which will result in a less representative ML community going forward.