And sometimes that works.

It's conceivable the issue here is one of habits and training - i.e. that one day we'll use such tools all the time, and have developed the right kind of instincts to code-review the suggestions that come up. Having a good feel for where tools are reliable, and where they need special attention is nothing new - and perhaps that just a question of practice here too.

Mind you, it might not be. It's sometimes harder to read code than to write it, so even if AI could write code just as good as the programmer, it's not obvious whether it's actually more efficient or in other ways better to merely select and review the AI-generated code vs. having a mental model and having a human write it. It may well be the reverse, at least for a while - that it's best to have a human use simple and predictable tools, and have an AI review the code afterwards to check for stupid gotchas. Or maybe AI needs to get a lot better before it's valuable for coding.

In practice I'm pretty unimpressed so far. It's really cool to see it work, but it's so unbelievably unreliable, that it sure comes across as a neat tech demo rather than something really useful. Then again, maybe I'm just not finding the best places to use it...

That just doesn't matter. There's so much low-hanging fuit; by the time we're at bunker fuel - who knows, maybe the renewable economies of scale will be so overwhelming that even solutions that sound absurd today will be economically viable. But if we manage to replace everything *but* bunker fuel we're still in a so much better spot than today that it's a moot point.

Basically: it doesn't really matter yet.

Though it's great we have some nuclear supply and certainly should keep currently running plants running, nuclear isn't even remotely cost competitive with plain old renewables such as wind and solar.

Nevertheless, given how dire our straits are, I'm all for saying yes to feasible nuclear power production increases - but you should be realistic and not expect miracles, here. Almost all of the fossil fuel replacements will need to be wind and solar, if only for purely economical and logistical reasons.

Well, prepare for true horror then, because if you're waiting for the hyper-rich to truly act sustainably, it'll never happen. Not entirely coincidentally, those same people [i]will[/i] find some way to personally shelter the wealth and wellbeing of those few individuals they really care about - themselves and their family, say.

Waiting for the hyper-rich or influential to stop flying (for instance) is an insane strategy for coping with climate change.

Pinning down where any given lie has cost decision making efficiency takes effort, and you're not always going to succeed, and you're often going to spend more effort years later nog just debunking the lies but then also trying to reconstruct how some specific knock on three-steps removed effect was harmful. If you look around, it's easy to see lies everywhere, all the time, and many are constantly harming people's decision making.

It matters where you place the burden of proof - and proving a lie is hard enough; additionally needing to pinpoint specific damage is (IMNSHO) too much.

Put another way: what's the social benefit exactly of having authoritative figures intentionally deceiving the public? It's not like anyone is forcing them to spout nonsense; and if that means we see a modicum of irrelevant self-censorship about whatever they ate yesterday: all the better.

But regardless of the morals of hypothetical examples: today's world is full of lies, and many are problematic. If we can tune any levers to even slightly reduce those, we should be trying that - especially when the cost (in the form of lost speech) is as low as it is with intentional deception. Nothing much of value to be lost there.

We protect those defamed from defamation; but we don't protect those deceived from deception. We should.

The problem with this line of reasoning is that it places the burden of proving harm from deception on those deceived, rather than having those with at the very least a duty of reasonable care having the burden of proving its harmlessness. To be clear: it's fine if to that burden to be light; but as is it's pretty much absent.

And that's a problem, because such harms are very diffuse, tricky, delayed, and otherwise hard to pin down, so it's easy to get away with.

At issue isn't really the ability to express an honest or reasonable opinion - but rather the absurdity of protecting the ability to not just make a mistake, but to intentionally deceive, especially when such fraud (which is what this really is) is perpetrated by those that appear to have authority and a duty to be truthful. And secondarily, why those that are deceptive have no real burden to at the very least correct their errors. Which they essentially never do, today.

In an entirely unrelated note it's true that it's hard to successfully prosecute (or sue for) many issues. However, that doesn't mean it's a pointless attempt - that threshold is significantly higher. After all, none of this is in any way rare; in fact I'd argue that prohibiting intentional deception in a few more cases is probably a lot more enforceable than some other laws, some of which haven't seen use in decades if not more. The point isn't whether you'll likely catch most infractions; its whether you'll catch some and discourage others, set against the risk of over-enforcement or misuse as a cudgel to silence critics.

The risk of over-enforcement depends on the details (as would in the US constitutionality), but similar if not even more serious issues plague defamation laws, yet those are still on the books everywhere. It certainly seems possible to balance those concerns, and I'd argue fraud is an even bigger problem than defamation.

5B is still too low. These fines need to be high enough to preemptively discourage even risking breaking the law. As is, given the limited legal resources available to regulators, a fine such as this isn't really leveragable more than once every few years, even if Google or some other corporate behemoth chooses to systematically flout the law until being called on it.

And that means the current system is broken; the costs of compliance (especially if they really succeed in crippling competition, or locking in consumers better!) may well be much higher than the fines, even if that fine is 5B once every 5 years (and that's being optimistic; they've never managed such a high throughput of major cases, ever).

As to the risk of actually damaging the company, which I'd say it *should* do: for a sense of scale, this is comparable to the volkswagen fine in magnitude. However, volkswagens revenue in the US was a tiny, tiny fraction of google's EU revenue, and it's market cap is less than 1 tenth of google's, *and* as a physical business it's much more sensitive to profit margins. Yet volkswagen is still an entirely viable business.

That suggests this fine should be probably more like 10 times what it is, even if that sum seems absurd in absolute terms. That still wouldn't actually put alphabet's long term profit even at risk, let alone represent a life-threatening situation for the company. I suspect the fine would need to be more like 100 times as large to really start damaging the company in the kind of ways that might be questionable; i.e. a 10x fine seems like it'd be a a more reasonable amount, assuming they desist promptly.

Alas, fines like this are capped by political negotiations that always smell like regulatory capture to me - they sound severe on paper, but are actually so low as to be structurally insufficient to actually pose a threat to any such large company; a great deal for behemoths, in other words.

This slap on the wrist probably would sting (if it sticks...) - but it's still slap on the wrist territory.

Let's not *entirely* forget that at the end of the day it's not up to the courts to think up the rules, there's an entirely branch of government to do just that. If we don't like these rules *systematically* then that's the only place a solid fix can come from.

Of course the HW costs aren't significantly covered by ads; certainly HW hypothetical costs from years ago, and not even to a significant extent HW costs today.

The original article's framing is deceptive (sigh). They're focusing on profits, which is odd enough to be plain wrong here. First of all, digital platforms have very low costs, so a large proportion of revenue can be profit - but the same doesn't hold for the physical TV - a much larger proportion of revenue will be due to the devices themselves. Secondly, the ad market is entirely dependent on the TV-market; so it's disingenuous to claim it's somehow making money distinct from the TV-market when in fact all that ad-based income is simply part of the TV business. That's akin to claiming a paid-subscription magazine makes most of its money from ads by accounting for the *costs* of the magazine on the subscription side of the business, and not the ad side of the business. In fact, just 10% of revenue was from platform+ (from vizio's numbers). It's not clear to me how much of that "platform+" business is actually ads, but in any case, the core model here isn't too different from an app-store; the numbers may change (though that wouldn't appear likely in the short to medium term). To the extent that "platform+" is indeed making profits it's also worth remembering that there's a trade-off there: increased profits in the form of ads *cost* profits on the hardware side: both because you need massive distribution to achieve advertising profits (so need to keep the HW margins razor slim), and secondly because having ads is at least somewhat of an anti-feature, discouraging buyers of the hardware at least slightly (i.e. - do you trust them with your private info? Anybody that picks a competitor over that trust impacts HW profits).

Engadget doesn't even mention the revenues, and chooses to emphasize the distinctiveness of the various streams of income: misleading reporting IMNSHO (ironically likely exactly the reporting vizio wants, because while you or I may distrust advertising businesses, for an investor they've been some of the more reliable gold mines - and they want to appear like a gold mine pre IPO, of course).

In essence this is just yet another company that owns a (small) platform and can therefore extract a bit of rent from it.

To be fair, a modern browser is considerably more complex than a classic OS, and for good reason. Seems pretty reasonably to have all kinds of fundamental stuff re-implemented specifically for it.

Heck, while Firefox may have lost the browser wars, just consider the fact they invented and maintained a revolutionary (depending on your perspective) programming language just for the browser, and at least technically it seems to have even kind of payed off, weirdly enough. And just how many JS compilers/JITs/baseline JITs/interpreters are there in a typical browser nowadays? Browsers have spawned multiple re-implementations of cryptographic algorithms too (also not exactly a top tip in normal practice). Then there's many a hand-rolled parser written in native code without parser generators that's exposed to actually malicious content all the time - is that also normal best practice? How about a custom bytecode transpiler aka WASM that runs untrusted code? Oh, and chrome decided to hey; let's just re-implement the whole network transport layer, because that's also a pretty typical move, a conventional best practice for sure. And because all this needs to look pretty, we've got custom constraint solvers (css/layout), a custom 3d engine that runs untrusted shaders in a wide variety of graphics cards (webGL), entirely new video codecs (AV1, VP9), audio codecs (opus), all of which again, run on untrusted input And to keep it all safe + manageable, it's best to make sure to run on only a tiny number of carefully selected possible configurations, with no extensibili-nononono, let's do the opposite of that.

The point being: if the argument is that some random program isn't worth breaking any molds for, then perhaps browsers aren't the best example.

Given the raging dumpster fire that is consequence-free discourse in US media, I'm not so sure that the standard in the US is actually better. Obviously, there are real issues with the UK approach, but it's kind of hard to ignore the massive real-world impact of online lies in the US; just look at 2020 (and really even many decades before that; successful disinformation campaigns have been pretty common).

Not saying we should emulate the UK, but nevertheless, the lack of consequences for spreading lies in the US isn't just some minor by-product of free speech rules we should clearly accept - it's a huge risk, and it kind of undermines the whole point of free speech, which is to protect our decision-making by retaining access to informed debate. So it's not OK to silence criticism but it is OK to drown out criticism in a wave of lies? The end effect is the same; those with access to those powerful tools can suppress critical voices and steer public opinion, just by another route. It's also worth remembering the fact that there are other gaping holes in the US implementation - it only protects from government censorship, which may have made sense when it was written (before modern incorporation was a thing), but clearly isn't working today. If the aim is to retain the ability to hold the powerful to account, it no longer suffices to constrain governments, it's now also necessary to constrain private entities that have grown so large and influential to have an absolutely crushing power imbalance with normal citizens, almost certainly more so that the government had any ability to when the constitution was drafted.

I think we should be more critical about the damage the US implementation of the concept of free speech does. The aim is noble; keep that! ...but the first amendment is really poor statute; it's both much too narrow in scope (only affecting some control, and only applying to the government) and self-defeating by construction (actually enabling control of discourse by protecting disinformation), and perhaps even misdirected (by focusing solely on protecting the speaker, when the fundamental aim necessarily includes the listener, who deserves protection from being manipulated at least as much as the speaker deserves to have a voice).

https://www.geekwire.com/2020/...

There is no data yet, but there is reason to be optimistic it'll reduce transmission. Let's hope that pans out!

The longer vaccine roll-out takes, the longer high numbers of viral particles are present in the population, merrily mutating. And that's particularly concerning when there's a vaccine around, because that represents selection pressure towards vaccine resistance.

At the moment we're limited by vaccine availability; so let's not panic. But once we get to the point were vaccine hesitancy actually affects the spread of sars-cov-2, it really is irresponsible for anyone that can get vaccinated to delay. Doing that will cause deaths, and if enough people do so, we risk the evolution of vaccine-resistant strain. Incidentally, that's also why we absolutely need to fund vaccines for poor countries. We really don't want any viral petri-dishes incubating covid 2.0 just because we're too cheap and callous - and think it's worth saving a few billion (some of the cheaper vaccines are just a few dollars a dose) at the risk of a new pandemic.

What other means of communication are likely as unfiltered and unmoderated? I mean, even twitter's fact-check labels are super-light-touch compared to all but the most subservient of media.

Notably, they didn't do this 4 years ago, and normally they try to keep up the facade that they treat users fairly, and as equally as possible. By comparison if a corporation is acquired, it's twitter handles are not reset; and no reset happened after the previous US election; and I can't find foreign elections that caused official account resets, so this treatmet is clearly unequal, and likely unfair (but to who?). Conversely, if this were a reasonable norm, it kind of raises questions about twitters whole model, because why is this distinction necessary? Probably because twitter profits off post-truth echo chambers, and fears that retaining the current followers might highlight that.