I think it's fair to say that requiring local administrative access to the device is out of scope, since you have already completely pwned the device in that case, which is what what you need to install a CA cert on any OSes.
But it seems likely that in the coming years, people will expect Apple's products to include the latest cutting edge AI (I don't mean useless AI shoved into every possible thing, I mean something closer to a useful Siri). Giving everyone else a 10+ year head start on you is not a good position to be in.
If you choose to outsource your AI to OpenAI/Anthropic/whomever, now you're beholden to another (risky), and for a critical feature of your ecosystem that your customers have grown accustomed to and to expect. And it's not just that they might jack up prices on you, but they can just... get acquired, or go bankrupt, or fall behind on model development...
At any point the interviewer could have clarified if they meant "at work" when they received an inappropriate answer. The fact they did not do this means they did not mean "at work," which makes sense because the questions they ask neither specify that nor are worded to make one believe they are work-related.
What would be the point of conducting an entire hour+ long interview where the candidate is only giving you irrelevant answers and you make no attempt to get them on track?
GP claimed “there doesn't exist any situation, in any plane in any conditions, where holding the stick back the entire time would be an appropriate input. Literally doesn't exist.” That's what I was replying to.
Because a court said it does not make it true. The captain made the correct choices. The first officer made inexplicably incorrect choices continuously; choices which nobody would ever make regardless of training, unless perhaps you had literally never flown a plane before.
At no point in time would the correct response to a stall warning be to pull the nose up. This is something taught well before you enter the cockpit of a commercial airliner. If you continue to pitch the nose up excessively, you will stall the aircraft. This is also something you learn on day 2 or 3 of flight school.
If the first officer had done literally nothing at all, 228 people would be alive.
The first officer wasn’t solo-flying a 172 with a six-pack cluster. They were flying an airplane that will refuse to stall in normal law. The plane had switched to alternate law 2, but the indication to the pilots was an autopilot disengage. That is bad design.
I also fault Airbus’s philosophy on countermanding inputs, though that warning is unambiguous and the pilots should have communicated about that. But when the damn “un-stallable” aircraft is yelling at you for putting the nose down, while also yelling at you for opposing inputs, you can’t not fault the plane.
I'd think we'd have outgrown things like claims of "unstallibility" given prior experience with ships claimed to be "unsinkable". There's what the marketing says, and then there's reality. You will never build a technical system incapable of being coaxed into an error state. Period.
Too many years as Quality Assurance grinds into you that when engineers claim perfection, you start questioning their assumptions, and the whole house of cards comes tumbling down shortly thereafter.
There's nothing wrong with writing CLI stuff in Node. It has interfaces to work with terminals built-in that aren't unlike what you'd find in any other language. I say this as someone who has worked with ncurses in C a decent amount (and there is a reason I try to avoid doing so anymore!)
React, however, I do find questionable, even having read about and understanding the idea behind Ink.
An EC2 instance with lots of cores like a c6i.32xlarge should do the trick, no? You could even pay for spot instances and just checkpoint frequently and copy the progress file to S3 when you get the interruption warning.
reply