Excellent! But: what he shows is that the entropy does not emerge from the implementation layer, but from the intentional layer. In other words, the bugs are not in the code but in the requirements. The universal development process begins with incoherent requirements and proceeds by guesswork until the compiler stops complaining. Anyone who has actually worked this process knows that pushback on buggy requirements is NOT tolerated. This is purely an organizational problem. How will he propose to mitigate it?
Mozart wrote for audiences who were only half paying attention. If that is all he had done -- and it was all that most of his peers did -- he would be forgotten. But at the same time he also wrote for audiences who were paying the closest possible attention. He is remembered for doing both. It is quite a trick, as you will see if you try it. Netflix do not even see the need for it, and therefore, their "works" will be forgotten.
Not quite as highbrow, but Pixar stuff, particularly the earlier movies, manage to have jokes that work for kids and their parents. It was much appreciated.
It already is. Every time they drop a new show, it's a hot topic for a week, maybe two, then it immediately falls out of the gestalt. No one brings up anything they've done in the future ever again. You barely ever hear anyone mention things like bird box.
oh I love the old shows that were written with two-level humor.
think foghorn leghorn with funny physical humor for the kids and subversive humor for the parents.
Sort of related -- I have friends who are immigrants to the US. They have a hard time with subtle types of humor, but some extra physical humor can sometimes let them have a good time anyway.
Who are the audience(s)? Who wants to hear which? They have been playing it both ways and will probably continue to. The moment to watch for will be if they start playing it only one way.
In our experience, it’s been smoother than expected.
Clean ownership pushes audits away from vendor trust questions and back onto internal controls. The only place we’ve seen more scrutiny is on the client IT teams that now own operation and governance.
My former employer was ordered by their auditors (who had permanent offices in the company's HQ building) to scrap homebrew systems that worked and adopt rhymes-with-SteepleCoughed, which did not and could not work, because the auditors' staff were trained on it.
I was peripherally involved in a failed effort to retrofit ACL onto another system, which ran on DB2 but had originally been coded on VSAM. After years of valiant resistance, they were forced to replace it with something not named Minimo, and to eat at least $60M of retraining costs, for the same reason.
He’s pretty clear in the fourth paragraph about one major issue. It only goes on more from there.
> Under Section 1201 of the DMCA, it's a felony to modify your own property in ways that the manufacturer disapproves of, even if your modifications accomplish some totally innocuous, legal, and socially beneficial goal. Not a little felony, either: DMCA 1201 provides for a five year sentence and a $500,000 fine for a first offense.
If it's all the same to you, I think I'd rather ask Musk how popular Nate Silver is. I won't learn anything either way, but Musk's reply might be more amusing.
"Humans must not blindly trust the output of AI systems."
Trust is blind. If it's not blind, it's not trust; it's something else, which may as well be called by its name, if it had one.
"Humans must remain fully responsible and accountable for consequences arising from the use of AI systems."
Today this is merely silly. We cannot enforce accountability among our own selves. That is why we want AI, so that we have one more thing to shift blame onto.
Like roaches: if you see one, you've got a million.
Also, here is one manufacturer, whom we never heard of, along with a citation of a handful of others, whom we also never heard of, who were afraid of reputational blowback from being blamed for spurious product. They are small. Probably they are right to fear. The major manufacturers, the ones we have heard of, do not share that fear.
Between my wife and I, we have now encountered at least three situations where generic drugs from one manufacturer were effective, from another manufacturer ineffective. Our doctor says that this is now a commonplace problem. Fewer and fewer pharmacies are willing to place special orders from specific manufacturers other than their supplier's default. And it was not so long ago that we learned that "turmeric" is now a euphemism for lead chromate. Think about what would be involved in enforcing quality control, and immediately it becomes obvious that it is not being and cannot be enforced.
reply