Don't forget the other end of the equation: has the time it takes for a customer to get their support issue effectively resolved decreased?

It may be the other way around: that the industry's idea of what "practical skills" means is changing faster than the universities' ability to keep up. By the time the Unis have adopted a technology, come up with a curriculum around it, found professors to teach it, and taught it to a graduating class of students, that technology is already considered obsolete and is no longer of much value to anyone looking to hire.

Dunno what the solution to that is, other than teaching the fundamentals and leaving it up to the students to apply them to technology stack du jour after they graduate.

Seems like everyone is already carrying a wearable in the form of their cell phone. If there's something to be gained here (which is debatable), they should provide it in the form of a free app, and/or add any necessary hardware to future cell phones, rather than trying to get everybody to remember to keep a second device charged and on their person 24/7 for the rest of their lives.

A big reason why health care is more expensive in the USA than in other nations is because the USA has a for-profit healthcare model. That means that the US healthcare consumer isn't only paying for actual healthcare, he is also paying for:

- "Increasing shareholder value" (read: funneling as much money as possible from sick people to Wall Street investment bros)
- Huge salaries for CEOs of healthcare and pharmaceutical companies
- 24/7 TV advertising of questionable drugs to people who aren't even remotely qualified to determine if they are appropriate or not
- Free lunches and treats for the staff of doctors' offices 5 days a week (because it gives the pharmaceutical rep a chance to promote their products to the doctors, and capitalize on the conflict-of-interest introduced by the doctor's satiated stomach)
- Huge numbers of full-time Congressional lobbyists (to help bend regulations towards what is more profitable and away from what helps patients)
- Large campaign donations every election (ditto)

If we switched to a non-profit model we'd be able to repurpose all that money towards providing health care. Then the USA could afford a single-payer health care system, like most other countries can, because our per-capita spending would be similar to theirs.

I'd think it wouldn't be too difficult to ship an EV with little or no charge in the battery. That doesn't make it fireproof, of course, but then again an ICE car with no gas in it is not fireproof either.

What Linux really needs is a native port of the MS-DOS prompt, complete the ability to select an arbitrary rectangle of text, copy-and-paste via context-menu, 8.3 filenames, 255-character path-length limits, and full support for executing all your favorite .BAT files.

Every bad actor is looking for a zero-day exploit to keep in their back pocket; why would you provide them with one for free, and assume they'll never use it on you?

Sounds like copium to me. Nuclear fission has lost the PR war, the economic war, and the tech war, and now you're hoping you can bring it back from the dead by slagging the competition with baseless hyperbole. Well, good luck with that; but it sure looks like that race has already been called. I still have hope for nuclear fusion, FWIW.

We know that the self-driving car companies have provided us with data that they say indicates that... but it's in their economic interest to convince us of that, and it's straightforward to lie with statistics if that is what you want to do. ... and even if we accept that "modern AI cars get into far fewer accidents than humans do", it's not clear that safety record extends to the LIDAR-shunning, camera-dependent Tesla implementation. Musk is making an aggressive bet that adequate safety can be achieved by a camera-only system; other companies think otherwise and have designed their systems accordingly. Time will tell who is correct.

Surely Tesla's supporters could debunk the whole thing by setting up their own tests that demonstrate FSD successfully detecting and avoiding the 'kids'? It's not terribly a difficult test to perform, and positive demonstration of correct functionality would be more convincing than just accusing other peoples' tests of being rigged/flawed/dishonest.

The exact values of "a bit" and "humans" are worth examining. Is it enough for them to just be better than the average human driver, or do they need to be as good as or better than the best human drivers? And if so, how much better is "enough better" to offset the social cost and complexity overhead of integrating the new technology into the road system?

... and yet, sporks are used only rarely, mostly by campers or at picnics, both of which are specialized niche use-cases where minimizing the amount of gear to transport justifies the necessary compromises in usability.

So, his analogy is exactly right. Most people don't want to use a spork, and will only use one in situations where access to a separate spoon and a fork isn't an easy option.

I can think of one reason -- supporting that would at least double the amount of QA they needed to do to validate each new release of either MacOS or iOS. That would be a pretty significant amount of overhead to support a configuration that most people didn't ask for and don't want.

OTOH if macOS was informally "ported" to the iPad by some non-Apple group, Apple might just look the other way and say "that's not supported by us, if you do it, no warranty, YMMV, good luck".

Don't forget the second-order effects of overruling your engineers' better judgement: your best talent gets frustrated with their work being sabotaged, and quit your company to work for one of your competitors instead.... leaving you with the less-talented engineers who are still willing/forced to put up with your bad ideas. Now you have bad ideas, implemented badly.

> Google Deepmind has a definition for AGI,

Thats not a definition, its just a set of subjective heuristics for measuring. And its not even as useful as the basic turing test, which is a much more concise yardstick.

Definitions of AI all seem to come down to "we'll know it when we see it" which is the exact same thing as saying "we have no idea what it is"