I shop around regularly. USAA isn't even remotely the cheapest option, and the handful of experiences I and family members have had suggest that they are just as shitty as any other insurance company when it comes to handling claims. I only stick with them because they are consistently within 20% of the cheapest option and I value having all of my banking and insurance handled in one place.
Comprehensive, collision and 500k liability through USAA, $120/month. Deductibles are as high as they will allow, which is $1k IIRC. I was recently forced to consider a move to Utah. Thankfully I didn't end up moving, but a quote for the exact same coverage in SLC was just $40/month.
I pay ~100/month per car for full coverage on two fairly new cars and $500k in liability from the auto policy plus a $1m umbrella policy. And that's in CA which is comically expensive. I find it very believable that you can get excellent liability only coverage on 3 cars for $100/month depending on the state and drivers.
As an American I would also like to know... My actual expenses for my low end luxury car are nowhere near that high, and I live in CA which has massively inflated car expenses in practically all fronts.
Registration: $600/year
Insurance: $1,500/year
Gas: ~$2,700/year (15,000 miles @ 30mpg @ $5.50/gal)
Loan: ~$3,500/year if I had borrowed the entire price of the car on a 60 month loan and then kept the car for 15 years as the GP stated
Maintenance: Certainly less than $1,000/year, much less in most years.
And in some states registration and insurance could literally be a third of what I pay. Gas could easily be half. I can't imagine anyone is paying $12,000/year for any non-luxury vehicle.
If two Waymo rides per day is covering your driving needs, there's no way you're spending $120/month on gas. Also, car insurance in CA is a huge ripoff, but I still don't pay anywhere near $180/month for very thorough coverage on my ~$40k car. Believable enough otherwise though.
I grew up with the fake stuff, and have lately done quite a bit of taste testing of the various grades of real stuff. Grade A golden is the only one with a distinct enough difference in flavor for me to care, and even then it's very much an acquired taste, not immediately and distinctly better. Given the stark price difference, my conclusion is that the maple syrup elitism is a silly hobby horse for people with too much money.
Way outside of my area of expertise, but a quick search suggests that exact numbers probably depend on exactly how you define the question, but it would be broadly reasonable to say that the balance is about 50/50 +/-5%, and trending towards the passive side over time.
I think it can be hard for anyone who hasn't done safety-critical software before to understand just how much testing goes into it. Even for non-human spaceflight, the stakes are so high, we have all of the testing you would expect as far as unit tests, system integration tests, end to end testing, but then we also have full human in the loop software sims, full hardware in the loop sims both at multiple levels of integration, and on and on. For every line of code I've written for a spacecraft, I'm 100% confident that if that line is responsible for a bug in flight, there are at least 10 other people who would share the blame with me.
Honestly, why? I can't speak for BO, but at NASA, we're all learning what the technology can and can't do just like everyone else. You can bet your ass that no one is vibe-coding any part of the rocket without thorough review of every line of code and thorough testing at multiple levels though.
California has mandated it for new construction for over a decade now. Of course old stock doesn't really ever turn over in California so it's entirely possible that we will never reach a point where a majority of SFH in California have sprinklers, but it's certainly possible to change for new homes.
If we're making that distinction, I think it would be more accurate to say that many people in the field appear to believe that these models are sapient, even though they are clearly not sentient.
Both anecdotally for myself and from what I'm reading in the news, it seems just as likely that AI usage has already largely peaked.
There was a lot of hype and exploration of capabilities, but models aren't evolving fast enough to keep that going, so I'm settling down into a familiarity with what an LLM can and can't do that means I am using them less overall that I was 6 months ago when I was throwing everything under the sun at it just to see what happened.
Without either new model breakthroughs or dramatically _lower_ costs, I will be very surprised if the ultimate market doesn't end up within an order of magnitude of where it is today.
Plenty of people have personal projects where the purpose is the end result. One small recent example of my own, I'm a long-time Garmin watch user. I struggled with learning all of the details of their custom language, runtime and SDK in the pre-LLM world. With agentic AI tools, it was dramatically easier to write a watch face that looked the way I wanted. From there, I very much did learn a lot and do a lot of re-writing and tweaking to make the code actually somewhat good, but the custom watch face was the point for me. I really couldn't care less about Garmin's esoteric tech stack. The only reason I bothered with everything after the vibe-coding step was to try to optimize power consumption where I could.
A key component of the GPL is the requirement that source of code of programs that use the GPL code be made available. Without IP laws, how would you achieve that goal of the GPL?
How does this work, concretely? Do you just upload your W2/1099s/etc and prompt it to "do my taxes"? I imagine it's not that easy, but filing my (trivial) taxes with any of the well known tax prep services is easy enough that the LLM version can't get very much more complex than "do my taxes" and still be an improvement.
Take this with a mountain of salt since I've not at all an expert, but with a little help from AI, it seems like the exemption lives in 2023/1670 [0]. The LLM claims that this and the regulation you link are interconnected, with the recital of 2023/1542 explicitly linking them.
>Unless your device complies to MIL-STD-810G CN1 and has the certification to back it up your product will be required to add user replaceable batteries
Can you provide your source for this? If nothing else, it's very surprising to me that an EU regulation uses a US standard as the baseline!
Edit: Having done a bit of reading on the standard, it also seems like the regulation needs quite a bit of detail if it really does rely on the MIL-STD, since the standard only defines test procedures, not pass/fail criteria?
You have to manually install hooks in your local repository. They aren't propagated as part of the repo. Git has intentionally made hooks require a very explicit opt-in.
I'm really curious how old you are. The great grand-dad part makes you think you're my age or a bit younger, but then I grew up very middle class and I don't know if I've ever even seen a black-and-white TV.
I'll imply those things. If you don't fit in the seat, you should have to buy two seats is a not very controversial opinion on the internet IMO. I think that opinion basically violates all of your "surely"s already.
Where do you draw the line? A 250lb person probably mostly fits in their seat still, but at some point a person is just physically going to take up two seats. Do you really think the airline should be responsible for flying them in business class (premium economy doesn't give you more width on most/all airlines)? Does it matter if their weight is due to a medical condition or just laziness? What if they're so big that even a first class seat won't contain them?
Except for "You make less money, often half.", which is a hell of a pill to swallow. As someone ~10 years into my software career, I'm pretty confident that even if I got laid off tomorrow and never found work as an engineer again, I'd still be better off now than if I had stuck with ME or EE as I originally planned.
I never played any games like that, but simply giving the agent a clear exit criteria and instructions to check the exit criteria every time it thinks it's done on a complex task was often enough to keep it chugging away for most of a day on a single prompt in my experience. Per-prompt pricing just isn't sustainable period, even if everyone is acting in good faith.
I'm curious about the opposite: Why would anyone use the CLI when, at least with Copilot, the VSCode plugin is super tightly integrated with VSCode, meaning the agent can see everything I can see. There's no mismatch in linter calls where I can see a lint in the ide that the agent can't find for example. I've had this problem even using CC in their VSCode extension, so I can't imagine it's not an issue in the CLI as well.
Which in turn owns Twitter. SpaceX is now a social media company in addition to a rocket company.
One theory I think Matt Levine posited, is that SpaceX will go public with dual-class stock that gives Elon control even with a minority ownership stake, and will subsequently buy Tesla, which doesn't have dual class stock, making SpaceX the singular "Elon Musk company", with him having operational control despite being public.