> The change here is that late night interviews are not "bona fied news"
Not always the case with Section 315, and late night and talk shows have been exempted in the past. The problem here is that this is on a case-by-case basis, and we have a particularly politically-charged executive agency.
> DoJ explicitly avoids JPEG images in the PDFs probably because they appreciate that JPEGs often contain identifiable information, such as EXIF, IPTC, or XMP metadata
Maybe I'm underestimating the issue at full, but isn't this a very lightweight problem to solve? Is converting the images to lower DPI formats/versions really any easier than just stripping the metadata? Surely the DOJ and similar justice agencies have been aware of and doing this for decades at this point, right?
This is speculation but generally rules like this follow some sort of incident. e.g. Someone responds to a FOI request and accidentally discloses more information than desired due to metadata. So a blanket rule is instituted not to use a particular format.
Image metadata is the wild west of structured text. The developer of the foremost tool for dealing with it (exiftool) has made 'remove metadata' feature but still disclaims that it is not able to remove everything.
"Fairly straightforward" is incorrect. Not an authority to describe in more detail, but the most tricky blocker I'm aware of are these proprietary "MakerNote" tags from camera manufacturers, which are (often undocumented) binary blobs. exiftool might not even know what's in there, let alone how to safely remove it without corrupting the file.
On the extreme end, simply decode the image and reencode it using an encoder that you have vetted to not include any metadata.
But I agree, presumably the image data part of the file is well and exhaustively defined. I would be very interested in counterexamples that have practical consequences.
Note that there will still be concerns about stenography and fingerprinting which would warrant such a disclaimer from the creator of a tool aimed at a nontechnical audience.
Maybe they know more than we do. It may be possible to tamper with files at a deeper level. I wonder if it is also possible to use some sort of tampered compression algorithm that could mark images much like printers do with paper.
Another guess is that perhaps the step is a part of a multi-step sanitation process, and the last step(s) perform the bitmap operation.
I'm not sure about computer image generation but you can (relatively) easily fingerprint images generated by digital cameras due to sensor defects. I'll bet there is a similar problem with PC image generation where even without the EXIF data there is probably still too much side channel data leakage.
I think there's a scenario where that's true: one where the head of your company is collaborative and deferential to expertise.
There's another scenario, though: one where the head of your company is a bull in a China shop, whose successes have come almost exclusively through a Barnum-esque scheme of cascading bravado and marketing genius without much expertise, but a marvelous ability to sell any idea purely via unearned gravitas.
The former is less sexy: I've compiled loads of talented people, and we're going to solve very hard problems, even some that seem impossible.
The latter is very sexy: I'm a genius and we're going to accomplish the impossible in one year via sheer force of my grand will. And even if it doesn't actually happen, I'll sell you on the next vision.
It seems like you’re ascribing to Elon some kind of magic, where you feel he’s breaking the rules of what should be allowed in order to achieve success. Is it impossible you simply don’t understand how what he does works?
So your hypothesis is Elon's domineering personality creates a culture of terrified silence where everybody wants to revolt but Elon is simply too powerful and they have no choice - and this extends to customers, sales and even technology - reality itself bends to the will of mighty Elon? And that's ... unfair?
Ok, I'm sorry, I'll try again. Seems I've missed your point a few times now, may be projecting my own perspecitve there.
So...it's not that you don't understand how what Elon does works, you do understand it, and your descrpition of him is accurate, you just seem to think it's unfair that it does work? "unearned gravitas" "w/o much expertise" "sheer force of grand will"
So you're saying Elon isn't a deferential technically-talented leader, he's wilful and a marketer, who you feel constantly changes course, and so maybe the people who work for him are not as aligned as I believe with what he's doing?
No, that's not true - the change was that you could only install software from verified developers, not only from the app store, and now they've partially walked that back too and "are building a new advanced flow that allows experienced users to accept the risks of installing software that isn't verified." ( https://android-developers.googleblog.com/2025/11/android-de... )
I will accept the Chief Emergency Shutoff Activator Officer role; my required base comp is $25M. But believe me, nobody can trip over cables or run multiple microwaves simultaneously like I can.
> But I mean, you should be able to be an odd duck and also be a director of a brand, as long as you behaviour isn’t hurting anyone…
I think the canonical example here is the Kirn Corksniffer[1] which could have been avoided with some foresight and subsequently a quick apology, but ego can sometimes get in the way.
Behringer definitely made bad products for a very long time, and while I appreciate the increase in quality the synth recreations don't really blow my mind. They're low cost, they're hardware, but almost all of them fall short of the originals.
> They're low cost, they're hardware, but almost all of them fall short of the originals.
As with everything Behringer, a lot of this varies from product to product. And it also depends which axes you measure along and which you consider important.
Like, for example, I don’t think anyone who’d done their research could seriously suggest the TD-3 and its variants aren’t as good as an original 303. In some ways they’re better: MIDI support, for example, and the MO - which is the second one I bought - implements the Devilfish mods that were popular on original units. But sound-wise, they’re as similar to a real 303 as the tonal differences between two original 303s. And hardware wise, well they’re cheap plastic boxes just like the originals so Behringer have certainly nailed the feel of using a 303.
On the other hand, something like the Poly D, does have some differences to the Minimoog. Again, it has some stuff that the OG doesn’t: an additional oscillator, paraphony, MIDI, a sequencer and arpeggiator.
Soundwise it’s very close but my sense is the filter doesn’t have quite the same hollow but fat character of the Moog filters I’m most familiar with (Moog One so not directly compatible). But it’s close and I’m not sure in a mix anyone would notice you were using a Poly D rather than a Minimoog.
And then you can find areas where corners have been cut: knobs and switchgear of the Poly D are solid and satisfying to use, but the keyboard is absolutely meh. It’s functional but it feels (and is) cheap. Not a patch on the OG.
Not always the case with Section 315, and late night and talk shows have been exempted in the past. The problem here is that this is on a case-by-case basis, and we have a particularly politically-charged executive agency.
reply