"It would be like a computer hacker saying that it's impossible to get a security vulnerability from buffer overflow if you use a language that always manages m

"It would be like a computer hacker saying that it's impossible to get a security vulnerability from buffer overflow if you use a language that always manages memory, and then that managed language suffers a vuln from memory corruption instead. Laymen three years later would only remember "impossible to get security vulnerability from managed memory"..." Ehh. The flip side of this is that most of the "experts" use this kind of language to hide their uncertainties, while also overplaying their hand. It wouldn't be unreasonable for the layman to remember the phrase "impossible to get a security vulnerability from managed memory", because that's the strongest part of the statement -- the part of the statement that's intended to sell a technology. People don't remember the long strings of caveats that come later, nor should they. My point still stands: we aren't very good at predicting failure cases, so it doesn't serve us well to focus on the caveats of a statement that begins with "this can never happen". Any statement that begins with an absolute like that is almost assuredly incorrect.