AI is stupid, but it can track details and at least provide a hint.

On the "make stuff that works" side, I've had the experience like yours. If someone lets it go too far, then it's a headache and it's easier to roll it back. If anything a bit painful as it's "empowered" some of the worst to make my day harder, being confidently incorrect all the time.

On the "find stuff that's broken" side, well, humans don't have the attention span and the AI techniques are catching little but critical mistakes by humans. In my codebase that has had regular security reviews for over a decade, including a few outside consulting companies, this year marked the first year that the teams had LLM at their disposal. It found two security issues that no one had noticed, one of which had been there almost essentially from the beginning. Admittedly, neither were exactly world ending (both required attacker to log in with admin privileges and the things you could do were a bit constrained), but they were real and undesirable. One of them kind of missed the point and although it mis-characterized the behavior, it put a light on an area where a human could actually think through and sort out the real issue, and how the flawed approach applied more broadly than the AI identified.

So AI as a review tool can work, it has a significant amount of false positives and misses stuff of course, but it can either directly catch or inspire human attention on a sketchy area.

The security team said that it was actually quite remarkable they only found two vulnerabilities that both required admin access, that most projects they dealt with ended up with 4-5 vulnerabilities exposed to either unprivileged users or even unauthenticated access. This is the sort of code the world is mostly built on.

Presumably, this is the whole point of these contracts, to hold the customers accountable for Micron making memory for them when the broader market is not necessarily looking for that particular memory.

Now I don't see how this can work out in one of the more well documented ones. OpenAI had at least a trillion dollars of these sorts of purchasing commitments, and even pretty bullish assessments don't support their ability to make that much purchase. So I do anticipate the market failing to make their minimum purchase commitments. I'm presuming there are penalties in these agreements and so they probably get money for nothing unless it gets so bad that OpenAI goes actually bankrupt.

It uses a screen as a gauge cluster, and puts the backup camera feed there.

They don't presume any speaker setup either, which would be a pretty key expectation of android auto in general.

They evidently will support a double-din mount of whatever you want, so you could add that without much issue.

Presuming reasonable access to channels to cable and mount speakers, I'm a huge fan of this facet of things.

I would like to see a couple of integration points, steering wheel controls and EV battery state/range estimates fed to the navigation like you get with other EVs and android auto, but no need to be as heavy handed as other platforms.

Unfortunately for my situation, it would lose out handily if a competitor had a midgate. I rarely need a substantial bed and usually need the seating, but it looks like it won't be easy to free up your bed if you have the seats in. Also subjecting the rear passengers to coming in by the front door.

If slate had a 4 door pickup with midgate to get longer bed, and maybe 20 inches or so longer to have that same bed even with rear seats available, I'd strongly consider it.

You do know that if the immune system got confused about virus versus cells hijacked to replicate viral material, we would have been dead pretty much the first second you got exposed to a virus? This is what a virus does. The difference is the virus produces material that will re-infect cells, while the mRNA resultant material is not self-replicating.

You also know that we had *billions* of people use mRNA vacinne and autoimmune problems weren't even vaguely a concern, and it was effective?

They say two things, one is that folks don't know how to deal with a released person:
"Their case manager may need to consult a dozen or more paper files or databases to learn whether they were convicted of a violent offense, if they require mental-health medications, if they can stay with family or need housing, and which vocational and educational programs they may have taken, among other factors."

The other:
"Not only would a more-efficient system help released inmates get the support they need, it could highlight who is likely to offend again after release."

So sounds like making more informed parole decisions?

Looking up a nuclear plant near me, the sustained power output is about a gigawatt.

Roughly looking up peak theoretical solar for a farm that could sit in the same footprint, it touches *maybe* 500 MW under impossibly ideal conditions.

Power density story is rough.

Further, the latitude causes some challenges seasonally for solar.

See, they found a way to get the support of the ones profitting from it.

The story makes pretty clear that they've been working this a long while, before at least the current hyped LLM was available.

To the extent "AI" might even play a role given their timeline, it was stuff that was pretty useless. People tried unleashing machine learning on these sorts of records and it just didn't do much.

Sounds like it's just a run at modernizing records keeping and access, which is fine.

Yes, but for a wealthy person, this is a much tinier fraction of their wealth.

The fact remains for someone under the threshold of the standard deduction, the property tax is something they have to pay that they cannot deduct, but a landlord could.

As I said, it's deductable without regard to the standard deduction. You can take the standard deduction *and* the property tax deduction but only if you are a landlord. I don't know how you decided to say "No you can't deduct that, it gets deducted"

A common person is somewhat potentially in posession of hundreds of thousands in house value. They are relatively less likely to have that much in stock except maybe their 401k, which is totally different.

Yes, there's sales tax. Ordinary income gets taxed that way on top of income tax. The leveraging unrealized gains as a loan is the most famous loophole, re-upping through re-borrowing at payoff and juggling that until death where there's a much more favorable estate tax.

The counterpoint is that the valuation seems to be a fiction when it could represent a liability, and a real thing when they want to, say, take out a loan against it.

It's awfully convenient that it is selectively fictional.

Note that for more humble "wealth", folks are taxed. If you own the house you live in, even if you are not using it as a financial instrument but just a place to live, you get taxed on the unrealized "value" of the house. I don't get to say the market value of my house is a fiction since I'm not selling it.

So we have a double standard, rich people wealth is selectively fictional with respect to tax burden, common person wealth is very much considered real for taxation purposes.

Even wilder, if you live in your house, your property tax is subject to the standard deduction, which means folks generally don't get a deduction for it. If you own a house that you rent out to someone else, the property tax you pay is not subject to the deductible, and you can deduct it. The tax system rewards landlords more than homeowners.

It seems that either you assess a property tax on net worth analogous to what is imposed on common folk or at *least* tax loans against such assets that have nothing to do with paying for that asset.

It's cheaper in the short term to get going. Methods without the lock in tend to be a bit more expensive up front. Most businesses are looking only at the current quarter without regard for the longer term. Buying is cheaper than renting only after some period of time, and that period of time is nearly always longer than a quarter.

Also, they have established their way as 'the' way for any *serious* business to do things. So management doesn't want to be seen as failing to do 'the' thing.

While true, you trade the VMware problem for a vendor that actively wants you to stop using your preferred product and start renting from azure instead.

Unless you are all in on azure, Microsoft is a dubious choice as a virtualization solution.

That device is $200. A random household that grabs a roku stick from walmart for $20 is not the same target demographic.

Ignoring the fact that a large share of smart TVs use Roku as their OS as well. So those random TVs are either Google, Roku, or some vendor specific thing.

Management has been pretty adamant that we are at the point where a developer should *never* look at source code, and work purely on prose. So some dutiful people have taken it to that extreme.

So they tell me they spend hours in planning, reviewing the resultant generated plan to give themselves confidence that the resultant code should make sense, then let it chew on making test cases, making code, running the tests, deciding whether the test cases were bad or the code was bad when failures happen, repeating until it has satisfied itself and then it lands in the builds that the still-human QA get a pass at. Then the human QA person often basically says "WTF, this is broken as hell" and the developer starts over incorporating the feedback from the tester into the flow.

Now the folks that discard that mandate and use CodeGen for code completion, more curated prompting, they get respectable speedups, but no hope for management to just toss those long term if they still need to curate the CodeGen. So management *really* wants the narative to work for the case above. They have raised the question if the real problem is the human QA, it passes all the CodeGen test cases and it's own code review, so maybe the human QA is just raising a stink to look relevant... Doesn't help that management understands neither the customer, software, or how to develop code, so they base things largely on wishful thinking on what the cheapest and most convenient answer would be that supports them getting big bonuses.

While AI might not be to blame yet for those figures, it is well within the era of TikTok where material that doesn't hit within 30 seconds, the audience is out.