At this point I'd be more worried about the government giving me a heart attack in my sleep.

They can't and shouldn't be trusted with this kind of power.

Let's be realistic here. The government isn't going to give you a heart attack in your sleep.

The government is going to give millions of people heart attacks, all at the same time, at 5 pm Eastern time on a weekday.

I cite the recent bonehead maneuver, wherein someone--likely backed by a foreign government--managed to access the data for every last person cleared to handle U.S. government secrets, which necessarily includes any embarrassing datum which could be used by foreign governments to apply leverage.

This is Murphy's Law in action. If a catastrophe is possible, it will eventually happen. Combine with Acton's Law. Power corrupts; absolute power corrupts absolutely. The result suggests that, for the sake of sane risk management, government should be structured such that no person working in it (or hijacking its infrastructure) can single-handedly do more than one nuclear bomb's worth of damage.

This means the fundamental principles that have to be engineered in are robustness, trustworthiness, and voluntary cooperation. Built-in backdoors are completely antithetical to all three.

> Do we also have nightmares about a hacker stealing a government's back door key and giving us a heart attack in our sleep?

While hackers interfering with embedded devices in people is worrying I'd be more frightened about bugs in the devices from sloppy minimum cost systems.

While not awesome current stuff is ok(ish) largely because it's a relatively small market so the people in it are better than the average dev.

I've been programming a long time, I'm a reasonably talented developer and I wouldn't touch that stuff at all.

Have you seen Ghost in the Shell? I'm don't more integrated computers are the answer.

But on the encryption front, the more the better. If everyone adds complexity then 'they' have to discriminate more. Which is good for everyone.

If you're going to weigh GITS as fictional evidence, you should at least weigh all other fictional evidence. Though you don't end up with much, except what knowing what people think about these topics.

I have seen Ghost in the Shell, although I'm not clear at what you're getting at.

One thing that stands out in my mind is the ability for someone to make himself invisible or unidentifiable to you by hacking your eyes. If someone can take control of your sensory inputs, they can put you into the "dark dream" to test, model, and eventually control your behavior.

When your eyes are hackable, you cannot take the VR headset off. Even closing your eyelids won't help.

Would you want anyone besides yourself having authorized access to your sensory bus? Would it make you feel any better if the person doing it had a valid court order?

  Would you want anyone besides yourself having
  authorized access to your sensory bus?

Banks' Culture novels touch on similar topics. In Excession, a Culture warship comments how the brain-computer interface (in-universe called a "neural lace") is the most effective torture device ever devised. And in Surface Detail, a firefight turns on the fact that Culture warships write their own completely-customized operating systems, with this heterogeneity making hacking attempts more difficult and consequently less successful.

GitS illustrates the danger of humans integrating with computers too closely ex. Cyberbrain hacking, stand alone complex, etc https://en.wikipedia.org/wiki/Philosophy_of_Ghost_in_the_She...

It's fictional but there's a lot of truth in no matter how important or shielded a target is, it is always vulnerable to attack.