Except that the word, "recall", doesn't have two different meanings. It literally means the same thing here as it does for a recall wherein a Jeep has to have its shifter replaced / tweaked / whatever.
Just so happens that the latter can't be fixed via software, because it's not a software-based safety failure.
The former is fixable via software update, because it is a software-based safety failure.
Both, however, are safety failures. Both need to be fixed. Both situations incurred regulatory requirements to notify customers and provide a timely fix at no cost to affected customers, whether within or without the vehicle's warranty.
In short: both are recalls, and recall means the same thing in both cases: notify, fix, or face regulatory consequences.
I too enjoy putting "quotes" around a word to make it seem like it means the thing I want to pretend it means, instead of engaging honestly with the actual situation. :-)
It should when it's required by the NHTSA and comes along with mandatory reporting, notification, and deadline obligations, along with optional free on-site execution of the fix.
That's what makes it a recall. A software update is the mechanism, the recall is the "Tesla did something wrong and is being held accountable" part.
I think you make a good point, benchmarks and metrics are indeed a better proxy for performance. Seems worth pointing out that, while "nowhere near half in [your] experience" are completely wrong, I don't take your word for it either. :-)
The trouble in my view is that the only way to know that the answers you're getting are accurate and not misleading is to study up on the answers elsewhere - which is a great habit to nurture, but is also precisely why these tools tend toward uselessness in their "general AI" bids. If I can't know how the answer was built, or how good that answer is, there's no point asking it - I'll just do my own reading and apply appropriate discernment as I go.
To be fair, hardly anyone does this today, nor did they before LLM-based chat bots... So it's a moot point, because society is largely doomed anyway. But a moot point can still be a valid one.
I also think the author makes a good point that we frequently confuse performance for competence. "It does a really good job at <X>!... or at least does a damn fine job of mimicking someone who acts like they do a really good job at <X>!"
By way of analogy, consider Elon Musk - by all appearances, he's a genius and is saving humanity - but by dint of his narcissism and largely smooth-brained approach to... well... everything... he's running all of us into an earlier planet-size grave than is necessary. His performance is fantastic, his competence is nonexistent.
Heh definitely an eyebrow-raising comment!
On a second read of it though, I took it more as "because minimizing the code written is a pattern observed in the wild, AlphaCode mimics it as that's what it learned from."
I was tempted to just post "because f** you, that's why" to be funny, but then it just felt like a quick path to the banhammer. :-)
Seriously though, I'm beginning to believe that this is the path of all general discussion platforms - jerks and morons infiltrate, and the rest of us fall for it.
Key difference - Uber didn't break the law while wiping out entire life savings volumes.
I understand that's a much broader-scoped situation than mentioned in the article (who cares about one country getting certain regs enforced), but the point stands - coinbase is a literally Ponzi scheme, and comparing it to other "eventually became legal" business models seems disingenuous.
"This is the value of cash. People can choose how much they want to trust 3rd parties and how much personal responsibility they want to take on. Some people may always choose a centralized mechanism for managing their assets, some will choose to manage their assets completely on their own, and others will decide on a case-by-case basis if they want a 3rd party involved."
"Just follow these 17 arcane steps and boom! Problem solved!" isn't a great flex.
That goes waayyy beyond "not wrapped up in a convenient user experience". Talking of finances is not so trivial a proposition as... Well, anything else that might have started our complicated on the internet.
"As such, if I stand up a new email server and start sending email to Gmail addresses, I am at the whims of Google deciding what email from me gets delivered."
This is the critical point. Adoption matters, and always ends up with consolidated power. It's just how things are, and how they stay, short of the severe intervention of dramatically easy-to-use tools for "click to host your own stuff".
I mean, sure, if you'd like to ignore the entire point of inherent or intrinsic value, then... go for it? Weird flex, but sure.
And no, it's not circular - "negative sum game" is literally baked into the concept of a purely speculative unbacked "asset". More money has to go in than can possibly come out, and there's precisely zilch to show for it.
Any place that accepts crypto is simply immediately selling it for fiat, and making the customer pay more than they otherwise would in fees and headache.
Anyway, it's fun to be obtuse, I know, but you're pushing the limits here.
Sure - today, as the situation stands, I'd sell it in a heartbeat, because there is still a vast supply of greater fools. That doesn't mean it has intrinsic value, nor that it would be an ethical / moral choice. Just means that I have loose enough morals to be happy to take an idiot's money.
When suddenly it loses its popularity and/or supply of greater fools, it'll fall apart, and I'll have taken some poor sap's actual money, even if it was delayed / distanced / distributed by some thousands of layers.
Value is subjective in many cases, yes, but there is still intrinsic value in the device itself, if we have to boil it down. You missed that bit - the intrinsic part is literally the metals of which the device is made. Even failing that, it's useful as a weighted object. You can get down to the nuts and bolts of a thing with intrinsic value, such that even in the vent that society collapses and there's no coordinated economy to speak of, there is still SOMETHING for which the thing/service can be used. Cryptocurrencies literally stop existing in such an event, and become entirely worthless in even less catastrophic situations.
Money is an agreed upon tool for exchanging items/services of value - but it also a store of value, because it retains at least a semblance of stability. Cryptocurrencies are... well, not currencies, for starters, but they're also simply a proxy for "real" money (ie: fiat).
Where cryptocurrencies entirely fall apart is when the realization hits that they all fall into the greater fool category - once you run out of new marks, the scheme collapses, and all the money - note actual currency - ends up in the hands of a few early adopters and a couple extra lucky folks who timed it right. Negative sum games are not a good thing to prop up.
Just so happens that the latter can't be fixed via software, because it's not a software-based safety failure.
The former is fixable via software update, because it is a software-based safety failure.
Both, however, are safety failures. Both need to be fixed. Both situations incurred regulatory requirements to notify customers and provide a timely fix at no cost to affected customers, whether within or without the vehicle's warranty.
In short: both are recalls, and recall means the same thing in both cases: notify, fix, or face regulatory consequences.