How so? I haven't had a phone call I've answered in years. A phone is useless. An internet connection is useful.

There are plenty of discretionary services that require support. Do I need a phone? No. Do I need cable TV? No. Streaming services? Nope. Amazon? Nope. Can I shop in stores instead of online because the online CS sucks? Yes.

People absolutely will pay more for better service. There is a reason I bought a BMW. It was the service center in town. The Audi service center treated me like a replacable comodity. I switched housekeepers to one more expensive simply because the previous one was rude. I switched landscapers because I didn't want a truck with a politcal flag swinging around my yard.

I shop at Ace hardware over the big box stores because they have staff that answers questions.

Service might not matter to you, but it does matter to many of us.

That's fair for mandatory spending. For discretionary spending I think capitalism will weed out those unwilling to support their customers.

So sure, the single choice internet provider serving you can do that, but the place selling you jeans can't.

Altman told the crowd that certain job categories would be completely eliminated by AI advancement. "Some areas, again, I think just like totally, totally gone," he said, singling out customer support roles. "That's a category where I just say, you know what, when you call customer support, you're on target and AI, and that's fine." The OpenAI founder described the transformation of customer service as already complete, telling the Federal Reserve vice-chair for supervision, Michelle Bowman: "Now you call one of these things and AI answers. It's like a super-smart, capable person. There's no phone tree, there's no transfers. It can do everything that any customer support agent at that company could do. It does not make mistakes. It's very quick. You call once, the thing just happens, it's done."

Tell me you have never called customer support without telling me you never called customer support.

And to be clear, that 680 Watts would presumably be the sum of all the multiple servers involved in responding to the query, the network switches, the routers, etc. It sounds high to me, but it isn't entirely implausible.

No, I meant exactly what I said. If the query uses as much power as running a 40-Watt light bulb for 17 seconds, but it takes only about one second to respond to the query, then that's the equivalent of drawing 680 Watts during that one second.

The caveat, of course, is that if you screw up and accidentally miss a payment, you likely get hit hard with interest, so this can be risky. Also, if you can't pay it off at the end of the interest-free period because you didn't plan for that well enough, again, you get hurt badly. So it's a calculated risk.

By wireless, I of course meant cellular (and point-to-point wireless, to some extent), not LEO satellites. That said, the same limitations that make traditional wireless ISPs suck also affect Starlink; they're just not serving nearly as many subscribers, and their subscribers are far less mobile, and mount permanent antennas outdoors, all of which make a huge difference. But still, those limitations will eventually start to be a problem.

Starlink's main problem (and, indeed, the main problem with wireless ISPs in general) is the ability to scale to a large number of users per unit of area. There's only so much spectrum. At some point, as the number of customers increases, bandwidth per customer decreases.

Right now, with O(2M) users in the U.S., Starlink service is still good enough for most people, meeting the minimum legal bandwidth threshold for broadband (100 megabit down/ 20 megabit up) for downloads most of the time, in most places, but not always, and not in all places. And apparently it averages only about half the minimum for upload speeds, on average. Now ask yourself if it makes sense for the government to subsidize adding another 2 million people, which would likely mean halving the speed.

Also, paying to put in fiber objectively increases available bandwidth for decades, and possibly centuries. Even if paying money to subsidize Starlink results in an increase in the number of satellites launched, the satellites last only about five years, so unless the government is willing to spend that money repeatedly, any gains are likely to be very temporary.

Like I said, Starlink is great, and I love that it exists. It has a lot of uses, and it has the potential to revolutionize a lot of things, like Internet service on airplanes and cruise ships, Internet service in RVs, cellular phone service out in the middle of nowhere, and so on. And it can be good enough for basic Internet service right now, given the current subscriber base. It can probably handle a decent number of additional people in rural areas without causing too much trouble. But it can never realistically be a solution for the problem of poor urban neighborhoods having massively worse service than rich suburban neighborhoods, because it just can't handle enough customers per square mile, and that is unlikely to change in the near future. That makes it not a great choice if you're trying to figure out how to spend limited subsidy money, unless your only goal is to cover rural areas, and it probably doesn't make sense to do that, because they'll get decent Starlink service even without the government subsidizing it.

17 seconds * 40 watts = 680 watts for a one-second search query. Realistically, that's probably the power used by the entire computer during that second, and there are probably multiple threads on multiple processors, so that's probably off by a single-digit factor, but it doesn't seem entirely implausible.

No idea for the AI number.

Why would you only make the minimum payment on a credit card? You shouldn't charge more on a card than you can pay off at the end of the month. Their interest rates are usurious. If you need to borrow more money than that, take a personal loan.

And the whole point of BNPL is spreading out the charges enough that you can afford to pay that amount every month. If you're not able to do that, then you shouldn't have made the purchase.

For good reason, to be fair. Starlink is able to pay for itself, and doesn't need subsidies to provide service. Wireless ISPs are going to suck no matter what, and no amount of subsidization will make it not suck. If you want bang-for-the-buck, you want fiber, because that can keep being pushed to faster and faster limits as technology improves, without changing the fundamental medium. Right now, I think the state of the art over a single fiber is one terabit. So we have three orders of magnitude of growth potential without any changes other than to the hardware at the two ends of that fiber.

Contrast that with celluar technology, where pushing speeds to orders of magnitude more than we have now can only realistically be achieved by massively increasing the tower density and, as a result of having more towers, also massively increasing the cost of every future hardware upgrade going forwards.

Starlink is a neat party trick. It can help with a lot of things, like providing service where it isn't really feasible, providing service to your RV, providing cell service in the Mojave Desert, etc., but it can't realistically ever be the ISP for the entire country, because you can't realistically put that many birds in the sky.

So fiber is the only plausible solution that is forward-looking and provides room for future expansion. Everything else is just wasting money, frankly.

Meh. Part of the point of that program was to provide jobs with decent pay. That's not really so unreasonable, is it? The real question is why the cable companies aren't willing to spend the extra few bucks to hire union labor for running their cables, in exchange for government subsidies.

Actually, no, the real question is why local governments didn't put in bids to build out municipal fiber networks that they could lease in a nondiscriminatory fashion to the cable company, the phone company, a dozen mom-and-pop telcos, etc. to provide the actual service to customers. This approach works way better than letting large monopolies or oligopolies get more power.

And at that point, it's just corporate welfare, and serves no real purpose.

And yet that's literally the only thing the government legitimately should be spending money on in this space. Every attempt to do this through private business fails. Every single time. Municipal fiber works. If you do it right (read "underground"), fiber requires very little maintenance except when somebody digs up a line, and then they're on the hook for paying the repair cost.

No, that's not it at all. Lines underground, assuming they are correctly marked and are at an adequate depth, typically last for decades. Lines on poles get broken by ice, falling limbs, lightning strikes, etc. There's just a lot more maintenance when you hang wires on poles.

Pole rights is a trivially solvable problem. You just pass one touch make ready laws. The fact that you don't have these is prima facie evidence of regulatory capture, and you should elect better representatives next time around. But for the most part, unless you live on bedrock, putting lines on poles is probably the wrong thing to do, so it probably isn't worth bothering to fix it at this point.

A policy of setting goals for the industry and rating them based on how well they met those goals? Well, I can't say for sure whether it was working, but putting blinders on and saying things are going great sure can't work better than having actual data, that's for sure.

To a limited extent, sure. The problem is that as long as the FCC is a political football — as long as Republicans don't actually care whether the poor have access to acceptably fast Internet service — the industry will generally not care much what the rules are, believing that the next guy will just overturn them.

It would have been continued, were it not for the fact that the Republicans tend to pick people for the commission that are basically poster children for regulatory capture.

In this case, the policies amount to requiring broadband coverage to show that they cover every house in a region, not just the wealthy houses, and that they provide service that meets certain minimum criteria for speed to every house, not just the wealthy houses.

The industry doesn't like to do this. They'd rather spend upgrade money on wealthy households, where they can milk them for higher profit margins, and never upgrade the service to poorer areas, even if they're willing to pay the money, because not enough people in the poorer areas will pay extra for service.

And from a business perspective, that behavior is understandable. But from a public policy perspective, it is problematic, particularly when it results in poor neighborhoods being stuck with, for example, ADSL at five megabit while three blocks away, there's gigabit fiber for not a lot more money.

It's relying on LLMs being able to search the internet for things. It goes somethign like this:

LLM searches for something, ends up on a site.
Site has multiple links embedded in it with said DNS payloads.
DNS lookup occurs.
LLM 'reads' said payload.

It actually is that hard, sometimes.

Often, jobs have a culture which have become structured so that you must be responsive, if not 24/7, then at the least during your work hours, to IMs. Step away from your desk for 30m to eat lunch or whatever? People are going to start calling you in many of these (IMO toxic) environments.

And frankly, it's required for some jobs (like in support roles). You've got to be available and IM is used for coordinating on the ground.

I've told people I am simply not available on IM platforms on my phone, I won't even install them if I can avoid it. This has caused some backlash, admittedly, but it's sanity worth preserving. If it's important, think it out a bit more and send me an email.

There's no good solution for this, unfortunately, particularly when everyone's set on using Slack for everything.

Actually, by putting down solar panels as covers over aqueducts used for the water supply, they can effectively make electricity and make more water (by reducing evaporation) at the same time.