It also outperforms GPT-4 for domain-specific user utterances,...

I think this is key. That is, I suspect Apple is not saying "we've created something that can do everything GPT-4 can do, but do it on the device." I suspect Apple is saying "for a very limited domain--such as asking your iPhone to call someone, or when asking for directions, Apple's AI outperforms GPT-4.

Apple's AI won't be composing poems in the style of Edger Allen Poe about Penguins (my favorite thing to do with ChatGPT), but it may be able to have a very limited conversation about finding a fast food joint on your map, or asking how to track your missing package.

I agree.

For example, consider that presently ChatGPT and other LLMs need to be trained on ginormous data sets beyond what any human being could read in hundreds of years. Yet in many ways humans still do better than ChatGPT at a lot of logic problems and math problems, and yet we are trained on a very small fraction of the data sets used to train LLMs.

And at some level I can't help but think we're basically anthropomorphizing a parlor trick; a pattern matcher that is so good at predicting how words should go together it almost seems alive to us.

Notice the second graph with the huge jump in 2022 labeled "Consumer fossil fuel studies...".

Consumer subsidies generally fall into the category of giving consumers money to buy fossil fuels. At least in the United States, one of the largest such subsidies is the LIHEAP program: the "Low Income Home Energy Assistance Program."

In general we count anything which "encourages" the purchase of fossil fuels as a 'subsidy,' but by doing that, we implicitly suggest that we should be allowing a lot of poorer people to freeze in the winter rather than giving them "consumer subsidies."

Wasn't the whole point of banning TicTok because they were supposedly spying on behalf of the Chinese government? And now what do we do--get them to spy on behalf of the American government instead?!?

I seem to remember this episode on TV...

I honestly wonder how much came from the Tonga volcanic eruption last year which reportedly dumped a "massive amount of water vapor" into the atmosphere? According to that article, it claimed that it could "take 5-10 years to fully dissipate."

The paper The Hunga Tonga-Hunga Ha'apai Hydration of the Stratosphere suggested that the one eruption added 10% to the amount of water vapor that typically resides in the stratosphere, and may contribute to global warming as water vapor contributes to greenhouse warming.

The Supremes ruled that the President did not have the authority under existing law.

The answer, of course, is for Congress to pass a law.

Of course the problem is then 535 people would have to actually show their support (or lack thereof), rather than do all this at arms length--meaning you'd have Democrats and Republicans actually voting up or down (rather than just bitching on Twitter). Meaning they'd have to put their money where their mouths are.

For me the compromise is to turn off all the notifications that make social media into a sort of "slot machine" on my phone. I've also managed to turn off all notifications across most of my apps--and I've even put e-mail notifications on a 15 minute delay, and only from certain individuals. That way I'm only interrupted when I think it's important, rather than because the app thinks it's important I hear about someone's ice cream photos.

It's been decades since I've worked at a company that went out of its way to hire technical writers and embed them with the developers to help write and maintain the documentation. And for some reason or another companies I work for have moved away from hiring QA as well: the company I'm doing work for has no QA testers; we developers are expected to test our own code.

I'd like it if we could move back to the days where we hire technical writers and QA engineers and testers; we need to remember that the best quality documentation comes from people who know how to write, and write technical documents. Rather than assume developers are "jacks of all trades," and that somehow we can solve the documentation problem through inserting document markers in poorly maintained comments.

After all these decades, I'm amused we're still seeing the aftermath of the fall of the Soviet Union, when Kazakstan was just another state in the USSR, and not a separate nation-state.

Of course not all of it is fun and games; Kaliningrad (technically part of Russia but geographically separated from Russia) contains one of Russia's most important naval bases--and yet is separated from Russia by Poland and Lithuania via the Suwalki pass. And if Ukraine all goes pear-shaped for Russia, this may be one of a couple of significant hotspots in a larger NATO/EEU war.

In capitalism you don't lose everything simply because a business fails. The company is liquidated, and you end up with a proportion of the liquidated assets."

Absolutely. And if there is no money left after the creditors are paid, or if there is a debt for which insurance has to cover it--as the FDIC did in this case--then the investors get no money, as there's no money left over for the investors to receive.

Years ago I bought $1,000 in stocks in WebGrocer. And it went tits up. And I got nothing back in my investments.

I took a risk, my investment was not guaranteed, creditors in that case got pennies on the dollar and stockholders like myself got nothing.

This is EXACTLY how this shit works.

The thing is, the "requirement" to give two weeks notice that most employees commonly deal with only affects when you receive your final paycheck: on giving two weeks notice your employer is then obligated to cut you your final paycheck on your last day. But you can quit under the law with zero notice--though it means your former employer now has two weeks to cut your final check.

In other words, commonly speaking, the requirement to give notice before leaving is not a requirement; it's a courtesy, and the consequence is simply logistical, not punitive.

So I'm curious about JPMorgan's contract; what are they giving for that six months? Because they cannot force you to actually work for six months--after all, what are they going to do if you decide to show up to work and just sit there at your desk doing nothing? Fire you? And it seems they reasonably can't claim part of your salary is in compensation for this six month exit requirement; it's your salary. I'd be curious what the contract actually reads, because if it reads "you shall give us six months notice" but does not have a consequence tied to it, then it's like those "non-competes" lawyers have you sign in California: it's a meaningless bullshit threat that cannot be enforced. And if JPMorgan says "or else you have to pay us a $50,000 penalty"--(a) how is that not buying your indentured servitude contract? and (b) if you get a good enough job offer, maybe you'll say "yeah, sure; here' your $50,000, now go fuck off."

I'm always suspicious of news articles that report some bad thing happening in society--but which fails to link to the underlying data. In this case, the article links to an earlier news article on CBS's own news site talking about preparing for retirement, and links to a site that plans trips for women--but does not link to the Census Bureau site, despite the fact that all such reports are provided and are easy to link to.

So I went spelunking, because I'm always curious about such things.

The best I could suss out with a few minutes searching was this article: Who Has Retirement Accounts?, which, interestingly enough, limits the survey to participation in 401(k) (and similar defined-contribution plans), IRAs, and "defined-benefit" retirement plans--traditional pension plans. And yes, the savings rate is dismal, with some 40+% of boomers not having a retirement account at all.

But... the definition here is "owns a retirement account"--that is, a tax-exempt deferred savings or defined-benefit pension plan designed for retirement savings. It does not survey overall savings. Now, the number of people who have a "retirement savings account" that is not a tax-advantaged retirement account--say, they own a bunch of shares in E-Trade, or they have equity in a house they plan to tap at some point--are not being counted in this survey.

It could be the percentage of people who don't have a retirement account but who has other savings is minuscule. But the point is, we don't know. And the CBS News article linked at the top is singularly unhelpful.

Worse, I can't help but think that this article, giving scary statistics but without a link to the Census Bureau's data sets (all of which are easy to browse and even contain APIs so that you can build your own synthetic data sets if you're so inclined) is designed more to sway public opinion than to inform.

It may be the public opinion they're pushing--that so many people don't have a safety net that any attempts at restructuring the safety net of Social Security would be heartless and cruel--may be the correct one. But we don't know. And we can't know given the way the survey was performed, nor will the average person know to search the Census Bureau's own web site to find out more.

Stuff like this bothers me, because I can't help but think such articles are manipulative, rather than informative.

In my own case, I do have a SEP, and my wife has a Roth IRA. If we were to be surveyed about the amount of money we have in those retirement accounts--people doing the survey would hold us up as examples of two people who would be severely impacted by Social Security reform, because we don't have anywhere near enough money in those two accounts to even think of retiring. On the other hand, we also have two non-tax advantaged stock accounts (one with E-Trade, one with UBS) which accounts for 95% of our net wealth (not counting the equity in our house). Personally I think we're okay, even if our tax-advantaged savings is laughably small.

I can see manufacturers balking at the idea of having to design around standard battery compartments,...

(Looks at all the gadgets I own which take AA batteries...)

So while I applaud these efforts, I doubt that we'll see them integrated into more expensive electric cars...

It seems to me that for a lot of driving in urban areas--which represents quite a few passenger miles out there--having a smaller scooter or even a skateboard style scooter makes a lot more sense than a single person driving around in a 4-seater 3,000 lb vehicle.

Don't get me wrong: I am **NOT** the target market for this; I live in a home on the semi-rural edge of a suburban development of a modest sized city who uses my truck to transport my recumbent bike to the greenway and our kayaks to the lake. But here where I live, there are a lot of college-age students for whom this would be the perfect vehicle for running errands and getting around town.

Looking at the actual data from the survey it appears they only surveyed banks, asset managers, transaction services companies, and pension funds.

They did not seem to survey companies who are actually engaged in forestry projects, land development or in other areas of the economy where they actually deal with trees.

I can't help but think this is noting but advocacy, and advocacy targeting banks to get them to divest any investments they may have with paper and wood product companies who aren't somehow publishing the right greenwashing phrases on their own internal publications--regardless of the care they may actually be taking of the land under their control.

It's also worth looking at the actual questions asked--many of which may be important (such as indigenous rights), but which have nothing whatsoever to do with "deforestation." (For example, Indicator #6: Does the financial institution have a policy that requires clients/holdings to ensure their business operations and supply chains meet key labour standards?) It's sort of odd to me that we'd be holding banks responsible for the hiring practices or land acquisition practices of individual companies. And tying all this under the rubric "combating deforestation" seems a bit... dishonest.