At that point what you are describing is a theme park ride. It only works a handful of times though before people get bored of it and want something else.
Back pre digital I was once lucky enough to see Aliens on one of the private cinemas at Fox, and it was astounding. I think people underestimate how poorly operated most normal cinemas used to be, combined with maybe not the best prints etc.
I remember seeing In the Mood for Love on the big screen in my local arthouse cinema back around 2000. It was shot with analogue film and projected as such, and the sheer details of the textures were astounding. It's not a bad film on my 4k monitor, but I don't feel the same awe.
IIRC in the film era there was one "master" of the movie, it was duplicated multiple times to make negative versions of the film, then those negatives are used to make the positive copies that are sent to theaters. So you're watching something that has been copied at least twice.
Commercial theaters are all digital now. They don't even have film projectors anymore. Some independent or "revival" cinemas might still have them.
> The US and the UK spent vast amount of money chasing exactly your line of reasoning with nuclear warhead decoys. Chevaline is a culmination of the effort, and it's retired for 30 years.
OTOH if you built a successful decoy system that is exactly what you would want people to believe.
Visual SLAM on a rocket would be wild. The frame rates you'd need at those velocities are brutal though — feature tracking falls apart fast when your entire visual field is changing at hundreds of m/s. Drones are the sweet spot where camera-based nav really shines.
That's because for short term they are really good, for the longer term the drift is too large so you will need many layers of sensors with different characteristics.
The return of Advogato. If you weren't around for it, it had a certification system like what you describe, so the stuff on it was pretty good. After a while, spammers figured out that it had very high search engine placement because of its quality, and that pretty much ruined it. It's gone now.
Apparently my account on the site is/was now more than a quarter of a century old... Gonna try to avoid thinking on that too deeply. :D
There's been a non-zero number of occasions since that time where I've observed situations that mirror the trust-based challenges Advogato sought to solve.
It is perhaps telling that as prescient as Raph's work on trust metrics was he later moved on to the notoriously challenging realm of font rendering--presumably because it seemed more tractable. :D
I sometimes feel like a paid newsletter that's curated by users would be fun. I'd happily pay €5 a month for a weekly/daily digest where the comments are en par with HN.
A few paid and unpaid newsletters have quietly become very big. Traffic from them completely eclipses this place, and because everyone gets the email at once it is a really sudden and painful spike.
Most I have encountered (generally via referral tracking) are heavily curated centrally though, and not by users.
The risk is to build very good echo chambers. One shouldn’t have to read AI slop or despicable opinions during their free time, but some exposure to alternative respectable and not idiotic views should be part of the design.
No software can do everything,
Good software makes concessions.
More concretely the ideal software would require a volume and speed of user input that is unfeasible, and the size and complexity of such software would balloon such that development and maintenance costs would be unfeasible, furthermore the more features, the more customization users want, so customization is an additional dimension.
But sure, go make a system that reads my mind and decides whether I want to trust a USER to do SOMETHING.
Yeah, with who is doing space exploration being right up there. If it is us it isn’t going to be in our organic bodies, and this renders so much of it irrelevant. Wider society will likely pigeon hole their thinking on that next to concerns about the heat death of the universe, but for a lot of us it is disappointing.
I did wonder about what it would be like embodied as a space probe encountering an alien that had also gone through the same process. That is now the sort of scifi that appeals.
I'd be very interested in any recommendations in that vein. I've been really enjoying the themes of embodiment in the new Marathon, where your body is disposable, woven silk with unfamiliar organs, while your consciousness is totally owned by a corporation.
The fact this reflects the subject matter in question is an irony surely lost on the author as well.
Political tedium aside a major factor in the decline of scifi is we live in the future, only it is not the future people were being excited about. As many creatives put it they wanted machines to do the chores and them to do the art, not the other way around.
LLMs are still not good at structurally hard problems, and it is doubtful they ever will be absent some drastic extension. (Including continuous learning). In the mean time the trick is creating a framework where you can walk them through the exact stages you would to do it, only it goes way faster. The problem is many people stop at the first iteration that looks like it works and then move on, but you have to keep pushing in the same way you do with humans.
Bluntly though, if what you were doing was CRUD boilerplate then yeah it is going to just be a review fest now, but that kind of work always was just begging to be automated out one way or another.
reply