Black holes can evaporate via some weird physics (and Stephen Hawkings' pop explanation is apparently wrong even as a simplification and he knew it...), but normal physical processes do not apply.

A black hole can't break apart. Any energy you add to the black hole to attempt it would only make the black hole gain mass.

I'm guessing:

1) The cloud of material surrounding the black hole normalizes over time just like a planetary disk. At some point, pretty much everything that can intersect with the black hole already has, and only random collisions create new infalling material.

2) When the black hole does feed, it produces a lot of high energy activity just beyond the event horizon, which pushes material away before it can cross.

It's really nice to have what is basically the world's most awesome anti-aliasing / magical magnification application there is.

The problem is that you cannot extract more information than there was in the source data, so when you add detail you are literally adding it (not recovering it). When AI adds detail, it can do some amazing things... but it can also hallucinate or average things towards a blend of its relevant training data.

It's a cheap shortcut that is unnecessary for most people and for the people who really care... it's offensively inferior.

If they cared about bots, there are several obvious currently active bot campaigns right now that are actively supported by subreddits' moderators. You just have to have the right audience that the wrong people want to influence and sit back.

Reddit only cares about bots when it doesn't result in them getting a piece of the action, or worse, where it drives their own action away. When it profits them, they support the bot problem with deliberate indifference, and this has always been their core philosophy even for things like child porn - until it became front page news, they had a LOT of underage nudity forums and massive amounts of activity in them.

There isn't really a profitable way to ethically run a social media site. You still need too many professional human moderators to control the bad actors who will endlessly abuse your forum if it has enough traffic that you'll lose money trying to 'do it right'.

Trump is not just threatening Canada economically, he's actively taking steps to destroy Canada's auto sector.

Why Canada should tie itself to an American industry that is actively hostile to innovation and increases prices for Canadians, while the US itself is hostile to Canada is beyond me.

Buy Chinese... better, more appropriate options and you're not giving money to a country that wants to annex you. When the US sorts its shit out, the auto industry can try and win the market back.

If a map service wants to take payment to make a business show on their map, that's fine. Good, even.

It's when that payment makes the business show on the map regardless of whether it's the class of business you're looking for or not that it gets to be a problem.

I don't need to know where the nearest coffee shop is when I'm searching for a gas station, nor do I need to know where Applebee's is when I ask about McD's. Those additional markers make the map less useful because they make it less clear.

They're going to outsource the decision making to a trained AI, then they'll say they can't precisely predict the choices it will make in advance, but that they will be in alignment with the training data.

For me, the real question is "is it statistically better than humans?", and wherever the answer is "yes", then we have to live with the machines killing people once in a while because they'll kill fewer people than people would.

I think you'll find you lose significant legal protects when you make credible death threats and appear to be interested in following through immediately.

Self defense can be a valid defense for your target killing you, and you don't have to sign anything.

Edge cases are why human override is required until the computer is as smart and flexible as a human being.

If I have good reason to believe someone is trying to get into my car to kill me, I am going to try to get away, and I am completely morally fine with that meaning I run them over if that is my only viable option.

The robotaxi would effectively sacrifice me so as to protect my attacker.

No, just ego. Shatner has always been an egotistical ass, but he gets away with it somehow; you wear a Starfleet uniform for a while and you get a pass. Fans link their egos with their love of Star Trek and suddenly it's almost impossible for him to do any wrong in their eyes.

I mean, the guy probably watched his wife drown instead of helping - despite quite a long time hosting a show about emergencies - but nobody questions it. There's audio recordings of him when he finally called 911, and he's left her in the pool instead of even trying to get her face out of the water first.

What kind of ass calls Sir Elton John 'whatshisname' and claims to have improved one of his classic songs?

We have heard Shatner 'sing' before. We've seen the results of his 'writing'. We've seen him act. He has fans because of his hammy Star Trek turn as Captain Kirk, and I'm one of them... but even at 95 someone needs to tell him to stay in his lane and show respect to people far more talented than he is.

A week after the discussion's done, you post that barely coherent slop in reply to an off topic post, that I have to assume was due to a compulsive need to defend Trump.

Wow. That's just so... sad.

You still need a firewall to handle non-signatories, and laws banning circumvention... it's similar to customs handling trade, except it's electrons and smuggling is almost trivially easy.

Two people stand across a border. On one side, a man in a country where it's legal to sell guns, on the other, a man who wants to buy a gun in a country where they're prohibited.

If the guy who wants to buy leans over the border and takes a gun, that's his country's problem. It's only the other guy's problem if no money is left behind in the gun's place.

No online service is going to adhere to the rules of every country in the world. If your country doesn't like that, it better get working on a really good national firewall.

AI isn't, so while it can immitate intelligence with approximate regurgitations of training data, it can't apply knowledge in novel ways.

What happens when nobody wants to share code because it's immediately vacuumed up by a corporation as training dara when they do? AI falls behind proprietary code.

The next big thing in programming is going to be closed code that manages to be opaque to AI.