In the "The End to End-User Elections" was this quote: "Drones hovering outside office windows will hijack a Bluetooth mouse to silently install malware on systems to tally who is our next president." And I realized, that's just barely plausible.
Security risk is an artifact of specialization. If you build a complex business or system, the greater the degree of specialization in its constituent parts, the more security risk you need to manage between them. It's that simple.
The made up nonsense names for invented problems are basically meaningless woo at this point. We could autogenerate security risk names and descriptions with fridge magnet words with a reasonable degree of accuracy, because the underlying cause is the externalization of risk from the gaps created by the internal incentives of integrating specialized functions.
It's not that all security is necessarily bullshit, but threats and risks that don't acknowledge this inevitable reality are. It's functional integration risk. This is funny, but when a world salad ML script can generate plausible products, the field needs to pause and reflect on what it actually does.
I began laughing aloud about here...
> Any given cloud at any given time has at least 14 zero-day attacker-controlled lightbulbs (they make them blink SOS in Morse)
...and then I stopped laughing when I realized just how true that probably is.
Ah! yes, Clausewitz’s 10 dictums about war, of course. My favorite one is “War is merely the cohesion of 5G wireless to malfunction a nationwide digital.”.
Man, this Clausewitz guy sure knew a lot about wireless technology for someone living in the 18th century.
"Flaws and weaknesses involving the deserialization of untrusted data will be a major concern, particularly in enterprise applications, like GRINDR, used directly by vehicle-based systems."
hahah