bignose

but from his perspective, that actually makes sense You then proceed to give the perspective of things that are not human. Which, sure, I’ll grant you that these gigawealthy arseholes are so out of touch they barely consider the human perspective any more.

For decades, when the horrors of advertising surface in the news, I’ve been dipping into the articles that commercial advertising people write for each other. It’s such a weird alternate world these people have constructed. One where the concept of consent simply doesn’t exist; where the general public do absolutely nothing with their lives except exist within one “channel” or another; where the right to blast advertising at everyone is so fundamental it can’t even be seen, let alone questioned. Not really a surprise. But my word, the edifices of thinking that have been piled atop these assumptions are mind-boggling in their fantasy.

Yeah im just wondering why they tried it at all. Because they have investors demanding insane profit returns, on a business that currently is not profitable at all. So they’re casting about for any different business model that might bring in some profit. Most internet business models don’t, so they’re trying advertising. They have no new ideas.

KDE e.V. has [verified] that Framework had severed their (already very limited) ties with DHH and Omarchy. I’d love to know a source we can verify for this? Haven’t seen any verifiable announcement of this severing of ties.

The major selling point for these LLM services is that people don’t need to learn a strict formal query language to interact; the LLM consumes instructions and data all as an undistinguished stream, and statistically infers a plausible response with absolutely no regard to the input’s meaning. That is: despite the media reporting these as “injection attacks”, that term means nothing when all its input, every time, is treated as data and instruction simultaneously. This isn’t some special class of attack; it’s a fundamental designed-in flaw of the system. The correct way to ensure protection from these vulnerabilities is long established, from decades of experience. You establish a firm boundary: never treat the input data as instructions, but instead have a separate channel for extremely well formalised query instructions, and reject bad input on that channel. But of course that would kill the major appeal for most people who love these things, the fact they don’t need to learn any strict formal language and can just say anything at all and get some useful-looking response. Take that away, and you lose any hope that the masses will want to use these systems. And so the makers and promoters of these systems will never make the one change that could even feasibly allow safety from these attacks; they will never make any improvement to security that might reduce the apparent ease of use of these things. For as long as that remains, these systems will continue to inevitably have these exploits because the corporations won’t close the exploit surface on these systems. These are staggeringly insecure by design, and can’t be fixed without being completely replaced.