Well, I probably wouldn't use that private key for anything else, just in case the browser gets compromised by something capable of stealing it, but it's still a dramatic improvement over sending email in plaintext. Might even use a dedicated browser strictly for email - if my webmail provider gets compromised and serves me malware capable of extracting my key, I probably have bigger problems.
Opposition Party... I've heard of such a thing. That's one of those things we had back before the two-faced single party system took over, right?
Well, since none of the sensors will report "big object ahead", that shouldn't be a problem. They'll all report a 1- or 2-D map of quantized data, which the software will then analyze to make sense out of. So long as it continuously cross-compares data from different sensors and "panics" if there's a substantial discrepancy (camera sees something that lidar doesn't, etc) it should be okay.
No, that was *their* claim - inaccurate in my opinion because humans *don't* have particularly elastic tendons. Kangaroos though *do*, and are by far the most efficient running animals on the planet. Despite their dramatic vertical motion they're nearly as efficient as wheeled vehicles.
Right, no external energy is required to keep a pendulum moving, for a while - entropy will eventually win, obviously, but you don't have to introduce any new energy to get the pendulum to climb back up the opposite side of it's swing. It won't climb quite as high as it started, but the initial energy is mostly conserved.
You are the only one suggesting that pendulums swing forever - they don't. Never have, and never will (barring the creation of a perfect vacuum and a 100% frictionless pivot). You can fake it by adding a mechanism to introduce additional energy to counteract friction losses, but then you're no longer talking about a simple pendulum, you've created a much more sophisticated machine.
Well, It would seem reasonable to argue that if the car is driving itself in fully autonomous mode, then you are not operating it. Would you charge a passenger with distracted driving?
As for recognizing the system in action... that's a reasonable issue. Perhaps indicators could be installed - sort of like "student driver" signs on training cars. Switch on autonomous mode and the signs light up, letting everyone around you know that this car is not under human control. Be great for DUIs as well, especially assuming a "black box" that can prove you you haven't taken manual control of the vehicle in the last X hours.
Automakers have long dealt with lawsuits over their various poorly engineered "death traps" - it's just a cost of doing business, it's not going to send them to bankruptcy unless they're already barely holding on.
I agree though that semi-antonymous/driving assisted vehicles have some serious attention issues. I rather like the current batch coming out though - if it can operate fully autonomously on the highway where things are relatively simple (and assuming it can handle crisis situations), then you relegate the driver to passenger status during that leg of their journey. That works - just make sure they're alert and at the wheel before control is handed back and there should be minimal issues. It may still only be semi-autonomous in a total-journey sense, but when it can take over, it can do so completely. Sort of like having a buddy in the car that will handle the highway driving for you.
Anything that requires the drivenger to pay attention while the car does the work though - that's just a recipe for disaster.
Well, the sudden stop on impact should be a good indicator...
Honestly though, a lot of it will likely be active monitoring. Software doesn't go bad - either it's the version released from the factory (easily tested via hash-check when starting) which will behave as designed, or it's not. If it's not, the car should refuse to run it. If there's a problem with the cameras/lidar/etc. that too should give some clear telltales. If the signal quality is not within expected parameters, or contradictory information is received from different sensors, the car can refuse to switch to automatic mode.
Of course that doesn't *completely* eliminate the risk of unexpected malfunction - but you can never completely eliminate the risk from anything: so long as the malfunction risk is appreciably less than the risk of human error, its a net win. There may be some serious PR challenges after the first few inevitable accidents, but so long as the autonomous systems are demonstrably superior to your average driver I suspect that will blow over, though it may well slow adoption.
And how does he NOT want police/media attention on the purported target in either of those scenarios? In fact, those are *exactly* the sort of scenarios where a call to the main desk would help prevent a lot of wasted resources.
Agreed. *IF* it were a claimed hostage making the call. Which it wasn't. And even if it was, it seems extremely unlikely that the hostage would just happen to be using an anonymized phone service. Not impossible, but it should send up a major red flag.
In fact I'd love to see some statistics on the percentage of anonymous 911 calls that end up being legitimate. I suspect the number is extremely low - maybe not quite zero, but low enough that the default assumption should be that it's a prank call. Still can't ignore it, but it calls for a very different initial response.
Yes - nothing there about perpetual motion, simply a reference to pendulum motion - you don't have to add energy on every swing, it will continue to to rise and fall without any external power source (at least until entropy robs it of it's momentum - but if we have to state that explicitly in *every* conversation it's going to get really tedious. Should we add "under the influence of gravity" as well?)
An Ada exception is when a routine gets in trouble and says 'Beam me up, Scotty'.