I should qualify that – I don’t know for sure whether and which distros enable updates to run non-interactively. fwupd has the ability to do so and it’s billed as doing so on its github page, but that doesn’t mean that a distro has to actually take advantage of that. Could be that in a default configuration on a given distro, it only updates stuff next time you invoke it.

The only reason I’d guess that it might not run automatically is that some devices do not deal well with power loss during firmware updates, and I can imagine that a distro – which has no way of knowing when a user might start flipping power switches – might want more-conservative settings. Might be something like the last bit of distro installation, but they might not want to run during normal operation.

But yeah, I bet that Louis Rossman didn’t think of that either when he was talking about using USB connectivity to prevent firmware updates.

EDIT: I also vaguely remember reading something claiming that smart TVs from some manufacturer that are not connected to the Internet were using nearby smart TVs of the same brand and within WiFI range that can reach the Internet for Internet connectivity. Ordinarily, I’d say that that’s not generally an issue for most devices, but printers often do have wireless networking capability, so probably one more at least theoretical vector via which a printer might potentially reach the Internet. I have not read any claims of a printer doing this, though. I also don’t know whether-or-not those claims for the smart TVs were legitimate, but they are technically-possible to do, so…shrugs

If we reach a point where an AI can summarize a video to that degree and provide background summaries, I’d happily use it. I mean, I’d mark the origin, but nah, this is just me.

Background from me: Basically, a number of printers are sold using a razor-and-blades model The printer is cheap. The ink is expensive. This is done because for a number of products, humans have a bias towards a low up-front cost, don’t weight ongoing costs as much – happens with phone plans that come with an inexpensive phone but make up the money over time by being locked to a service that cost more, for example. So if a manufacturer can put a printer on a shelf that has a lower up-front cost, uses the razor-and-blades model, they get the sales, not the one next to them that has a high up-front cost but lower costs for consumables. Inkjet printers manufacturers had been increasingly-widely doing this for some years, with printers getting cheaper and ink being sold at increasingly-higher prices. Third-party ink manufacturers picked up on this and started selling ink at a much cheaper price. This dicked up the business model that printer manufacturers have, and printer manufacturers fired back by building authentication chips into their ink cartridges and similar.

For some time, this was pretty much entirely the province of inkjet printers. Getting a laser printer tended to avoid that. Brother is a prominent laser printer manufacturer that made printers that didn’t have restrictions being placed on them, so was often recommended as a way to avoid all this.

Rossman: What Rossman’s saying is that Brother has started doing this as well now. He gives some examples of firmware updates being pushed out to Internet-connected Brother printers to cause them to stop accepting third-party ink cartridges, as well as some other behavior that he considers anti-consumer. He had previously recommended Brother monotone laser printers as a way to avoid this [I had as well]. He made a wiki page listing all the things that they’re doing. He says that he doesn’t know of a type of printer to recommend now.

He then spent a while being licked by his cat, who he says likes the taste of his skin cream. A substantial portion of the video is his cat licking him.

Yeah, that can cover some cases (also, throwing data on a smartphone, which most people have and keep with them most of the time) but I think that for most people, electronic devices still aren’t a complete replacement for paper.

• Power. Paper just needs some kind of light in the environment.

• Shareability. Okay, there are schemes to let one transfer data from phone to phone, but it’s hard to compete with how intuitive and universal handing some paper to someone is.

• Battery. Just keeping the display on a phone or laptop, even if you aren’t far away from power, on to keep the page visible tends to consume power, and many devices can’t keep something visible all day. I’ll concede that eInk displays can cover some of that.

• Disposability. Paper is pretty cheap, and if a piece of paper gets soaked in water or whatever, it’s no big loss.

• Use of paper in the physical world. I can do things like create stencils on a sheet of paper and cut them out. It’s a device that lets a digital computer interact with the outside world beyond purely showing information.

We’re a lot closer to the paperless world than we were when I first started hearing the phrase “paperless office”, and a lot of documents never leave electronic form, but I still do occasionally want to use paper.

Professional writing was always fake.

I don’t even know what that means. You mean that professional authors use spell-checkers or something?

Strictly-speaking, in this case, it’s not the ability to be network-connected that’s at issue, but rather the ability to push updates to firmware.

I don’t know what type of computer you have it connected to, but Linux has a system that will automatically update firmware on USB-attached devices if the attached Linux computer is Internet-connected.

$ sudo fwupdtool get-devices

Will show you a list of managed devices.

I’m sure that Windows and MacOS have comparable schemes.

On Linux, I’m sure that you can blacklist a device for updates.

I’d guess that it’s possible to get one of those dedicated USB print servers. Those probably don’t support updating firmware on an attached printer. I might have some questions as to how much I’d trust a no-name one of those on my network itself, but…

Honestly, that’s not a terrible idea in general. Like, if you have an Internet-connected device, you have a hook onto your network that someone can exploit down the line, including – as Rossman points out – making it function differently than it did at the time of your purchase in ways that you may not like. And even if you trust the manufacturer, that doesn’t mean that someone cannot acquire them and then exploit that hook.

Kind of a problem with apps and other software too. Even open-source software, like the xz attack – the xz package itself was fine, but you had someone, probably a country, intentionally target and try to seize control of an open-source project to exploit the trust that the open-source project had built up. I understand that it’s also been a concern with even browser extensions.

The right to push updates to an Internet-connected device, unfortunately, has value. And there are people who will try to figure out ways to take advantage of that.

Ehhh. I rarely print anything, but I really don’t want to give up the ability to print things at any time I want and have them promptly available.

I had to laugh at this. At least in my use case, it’s printing out forms and documents that various levels of government needs and I am absolutely not talented enough to reproduce them by hand (also, my handwriting is not fantastic).

If we want to get pedantic, it is possible to get a pen plotter. There are fountain pen compatible pen plotters, and the whole fountain pen world has a healthy and mature third-party ink market.

Now, that’s not simply a drop-in replacement for a regular printer, starting with the fact that you need to use monoline fonts so that the plotter traces out what a hand would rather than filling it in, and that a plotter just can’t produce all the same stuff. The speed is going to be abysmal compared to a conventional printer for virtually any image. And I don’t think know if there’s anyone who has built one with a paper feed system (there are large-format pen plotters that can work with a continuous-feed roll of paper, but I don’t know if those can handle fountain pens. I don’t know of a fountain pen plotter that can just take a ream of A4 or US Letter pages and then handle those correctly).

But you can, strictly-speaking, have a computer create output that uses ink from the fountain pen world.

Ugh.

$100 mono laser printer

Well, you probably aren’t getting a $100 laser printer unless they’ve got a razor-and-blades model. I definitely paid more than $100 for the mono laser I have. I don’t know what printers out there are gonna be fine with third-party ink (or toner), but any that do are going to cost more, because they aren’t relying on ink sales to make the printer business viable.

He says that he doesn’t know what to recommend any more, now that Brother has started doing this too.

I understand that Epson has some inkjet printers that don’t use ink cartridges. You just pour more from a (cheap) bottle into the tank. Like, they can’t implement a lockout, and there are other manufacturers that sell ink for them.

kagis

“Ecotank”.

https://www.amazon.com/s?k=ecotank

But if you want those, they’re gonna cost more than printers that are using the razor-and-blades model and expecting to make their money on the ink.

https://epson.com/For-Home/Printers/Inkjet/c/h110

There’s a list of their home inkjet printers. Notice how the “EcoTank” ones cost more than the non-EcoTank ones.

Like, one way or another, the printer manufacturer is gonna make their money. Either it’s not razor-and-blades model, in which case the printer is gonna cost more but the ink is cheaper, or it’s razor-and-blades and you get a cheap printer but pay more in ink and the printer manufacturer will do everything they can to lock out anyone else from selling ink for the thing.

EDIT: I’d add that I am not personally a huge fan of inkjet printers unless one really needs what they can do, like printing photo-quality images, because they have so many more issues with ink handling than do lasers. I can have laser printer sit without powering on for five years, then turn it on, and it’ll come right up and work fine. Inkjet printers are prone to clogging problems.

I loath them because I don’t have pockets

This is especially an issue for women, who often have more form-fitting clothing that either doesn’t have pockets or have very small ones that don’t work for phones.

I think that the usual solution for “women carrying things” is that many are gonna carry a purse – if someone’s pre-menopause, they’re gonna need pads or tampons anyway, so can put it in there. Problem is that the phone breaks this. Even if women have a purse, women don’t always carry their purse all around the office or house or whatever, but don’t want to miss calls.

My mother got a fanny pack just for her phone (which isn’t even all that large).

At one point in the past, it used to be common for women to wear a bag on a belt accessible through a slit in their skirt.

https://pieceworkmagazine.com/a-brief-history-of-the-pocket-in-womens-fashion/

The first examples of pockets began to be inserted into men’s clothing at the end of the 1600s. Before this construction development, illustrations show that men used small pouches, which hung from a belt around the waist. These separate pouches could be concealed inside of a coat or tunic. The words pouch and pocket are related, through the Middle English/Northern French word pouche, originally describing a small bag.

For women, pockets remained an accessory that tied around the waist and was accessed through an opening in a skirt’s seam. The full skirts of the 1700s allowed these pockets to be easily hidden.

The shift by women to pants kinda killed that option.

I think that the solution is gonna be some women’s clothier figuring out how to make an appealing way of carrying a phone.

Lara Croft runs around with thigh holsters. That way, the carrying system is clearly distinct from the body, doesn’t mess with the body silhouette, which I assume is why women don’t want male-style large pocket, non-form-fitting clothing. So maybe something like that would work. Dunno how much of a chafing issue that is.

EDIT: Drop bags are kinda in the neighborhood of what I’m thinking of too, though I’m thinking lower-slung and smaller:

Just leave it plugged into the headphones, don’t even take it off. I mean, I have 1/4 inch audio hardware, and I’ve got 1/8 inch headphones that have a 1/8 inch to 1/4 inch adaptor that just lives on the end.

I totally understand people who want to use wired, TRS headphones. They’re inexpensive, widespread, aren’t going to become e-waste when their battery dies, aren’t going to become obsolete when radio protocols move on, are lightweight, don’t suffer from radio interference etc. I have a bunch of TRS headphones and like them. Only downside is that they need some power source if you want to do ANC, but it’s not like one has to have ANC.

But…I think that a lot of people are treating it as a “we live in a Bluetooth world or a wired headphones world, and which we do depends on whether there’s a TRS jack on the phone itself”.

I’d also add that if you have a USB-to-TRS device acting as your DAC, you can swap in others, aren’t stuck with the on-phone DAC. I had a phone that had an extremely obnoxious tendency to, when charging in the car, play noise back through the headphones jack (and thus to my car’s aux jack and through the speakers). Was fine on Bluetooth. Problem was that the manufacturer had failed to stick the proper filtering circuitry in the power supply for the DAC and was spewing noise from USB power into the audio output, probably because you couldn’t see a problem when the phone was running on battery and filtering circuitry for the DAC uses up space in the cramped confines of the phone. (In practice, USB power can be amazingly dirty – I was astonished watching some people with oscilloscopes look at the power lines on USB.) Anyway, the noise was appalling. If you use the built-in DAC, you can’t really change the thing out. With an external DAC, you can stick a reasonable one in.

I don’t know how the ones I linked to above perform. But I’m confident that if they are a problem, there are other DACs out there. Whereas with a built-in jack, you get the DAC that the phone manufacturer provides, and clearly some are willing to ship their phones with an inadequate DAC.

I’d kind of like to see someone set up a rig with intentionally-dirty USB power and a bunch of USB audio interfaces and USB-powered devices with an audio output and then see how much noise leaks through into the DAC’s output.

EDIT: I also had a (purely analog) audio mixer at one point that used USB power and also leaked audible – not as bad as my phone in the car – noise from the USB power source into the audio. Solved that by moving it from my computer’s USB output to a dedicated USB charger. I’m sure that there’s still leakage and if I were doing pro audio work with that hardware, I’d still be looking at it, but at least it isn’t easily-perceptible to me any more.

I also had an inexpensive USB audio interface that leaked a little audible noise into its output, one of these:

It wasn’t terrible — I used the thing for years — and on that, moving the USB cable around would adjust how much audible noise was making it out the DAC’s output, so it was definitely unfiltered noise coming in from USB power.

I think that it might be underappreciated how bad the DAC situation in home electronics is. I haven’t seen people trying to measure and quantify it. I have seen lots of people going to great lengths to measure frequency response on headphones, whether or not a digital data cable has (probably completely unnecessary) shielding, and worry about the encoding of their music and sometimes even its encoding for wireless transmission to headphones over Bluetooth. But “how much junk from the power source is leaking into the DAC’s output” seems to be a curiously un-measured area.

The judge wrote that he “does not aim to suggest that AI is inherently bad or that its use by lawyers should be forbidden,” and noted that he’s a vocal advocate for the use of technology in the legal profession. “Nevertheless, much like a chain saw or other useful [but] potentially dangerous tools, one must understand the tools they are using and use those tools with caution,” he wrote. “It should go without saying that any use of artificial intelligence must be consistent with counsel’s ethical and professional obligations. In other words, the use of artificial intelligence must be accompanied by the application of actual intelligence in its execution.”

I won’t even go that far. I can very much believe that you can build an AI capable of doing perfectly-reasonable legal arguments. Might be using technology that looks a lot different from what we have today, but whatever.

The problem is that the lawyer just started using a new technology to produce material that he didn’t even validate, without determining whether-or-not it actually worked for what he wanted to do in its current state, and where there was clearly available material showing that it was not in that state.

It’s as if a shipbuilder started using random new substance in its ship hull without actually conducting serious tests on it or even looking at consensus in the shipbuilding industry as to whether the material could fill that role. Meanwhile, the substance is slowly dissolving in water. Just slapped it in the hull and sold it to the customer.

EDIT: Hmm. Actually, I thought that the judge was saying that the lawyer needed to use AI-generated stuff in a human-guided role, but upon consideration, I may in fact be violently agreeing with the judge. “Actual intelligence” may simply refer to what I’m saying that the lawyer should have done.

I assume that the author specifically wants a smartphone.

3.5 jack.

They exist, but it’ll constrain your phone choices a lot.

I’d just get a USB-C-to-1/8"-TRS adapter. If you want to charge while playing, you can get one with passthrough.

Without passthrough:

https://www.amazon.com/Anker-Adapter-Female-Samsung-Devices/dp/B08Z3B5QL3

or with passthrough:

https://www.amazon.com/ZOOAUX-Headphone-Charging-Earphones-Compatible/dp/B094Z6149B

Can probably just leave the thing plugged into your headphones.

Why can’t we go back to small phones?

The iPhone SE is dead,

Is there any chance that you chose to lock yourself into a very small walled garden with a vendor who might make decisions about product that you might not agree with?

Apple is the only one making iOS phones, and Apple doesn’t seem interested in small devices anymore, so that door is shut.

Right. You stick yourself in that garden, you are gambling that the vendor is going to come out with the product that you want.

There are still a few niche companies working on smaller devices, like Unihertz, but those phones almost always have low-end hardware and limited software support.

Well, size is kind of a constraint on what hardware you can put in the thing.

If what you mean by “limited software support” is “apps are going to be optimized for the bulk of users and will probably feel small if the great bulk of users are using larger screens”, well…I mean, yeah.

The iPhone 3 SE you have:

4.7-inch (diagonal) widescreen LCD Multi‑Touch display with IPS technology

1334-by-750-pixel resolution at 326 ppi

Memory 4 GB LPDDR4X RAM

https://www.gsmarena.com/results.php3?nYearMin=2022&nRamMin=8000&fDisplayInchesMax=5.5

Let’s grab one from that list:

https://www.gsmarena.com/ulefone_armor_mini_20t_pro-13298.php

Size 4.7 inches, 53.3 cm2 (~63.1% screen-to-body ratio)

Same screen size as your phone.

Resolution 720 x 1600 pixels, 20:9 ratio (~373 ppi density)

30 pixels narrower, but 266 pixels taller than your phone.

8GB RAM

Twice the memory of your phone.

Can buy online in the US:

https://www.amazon.com/Ulefone-Armor-Mini-20T-Pro/dp/B0DJ74TQXT

And it was released October 2024, so it’s pretty new.

Now, you may not be able to get an iOS phone that fits your hardware wants, but them’s the breaks when you go with a platform that has only a single vendor making hardware for it.

You’re correct. Not all software will be able to usefully run under such a restriction.

But I think that a very substantial amount probably can.

Given that a lot of Python software I see already by convention runs in a venv, which is at least somewhat-isolated from the rest of the system…I wonder how much harder it would be to make it the norm for most Python software to run in an isolated sandbox, without broader filesystem access. Like, kinda follow an Android-like model, where there’s an application-private directory and permissions that mostly make the app keep to itself.

I do run some Python software in firejail. But it’d need to be a norm for how the software is distributed. Can’t require a bunch of technical work on the part of end users.

Of course, Altman is referring to chonky enterprise-grade GPUs like those used in the Nvidia DGX B200 and DGX H200 AI platforms—the latter of which OpenAI was the first to take delivery of last year.

You wouldn’t be using these for gaming (well, not of the 3D graphics sort).

They run in the tens of thousands of dollars each, as I recall.

Probably more correct to call them “parallel compute accelerator” cards than “GPUs”. I don’t think that they have a video out, even.

What they do have is a shit-ton of on-board RAM.

EDIT: Oh, apparently those are whole servers containing multiple GPUs.

https://www.trgdatacenters.com/resource/nvidia-dgx-buyers-guide-everything-you-need-to-know/

The NVIDIA DGX B200 is a physical server containing 8 Blackwell GPUs offering 1440GB RAM and 4TB system memory. It also includes 2 Intel CPUs and consumes 14.3kW power at max capacity.

For comparison, the most powerful electric space heater I have draws about a tenth that.

DGX H200 systems are currently available for $400,000 – $500,000. BasePOD and SuperPOD systems must be purchased directly from NVIDIA. There is a current waitlist for B200 DGX systems.