Use something that supports tags properly.
It lets you handle fuzzy boundaries way easier. If something’s both fantasy and sci fi? Give it both tags. A book on the real science implications of some fantasy magic system, using actual quantum physics models? No problem. Give it fiction and non-fiction, and science and fantasy.
Then you can filter by tags to get all the books that fit what you want.
What do you think about translating games with questionable quality? Should such efforts be encouraged or discouraged in the gaming community?
I wouldn’t say encourage it. It makes the most sense for the community willing to do the work to push for games that people will like.
But there’s no reason to discourage it. If that’s what they want to do, go for it.
My interpretation of what they’re saying is that it’s on their own servers in their own location that can only be accessed from specific access points.
Talking about networks as airgapped isn’t abnormal.
I’m not really arguing the merit, just answering how I’m reading the article.
The systems are airgapped and never exfiltrate information so that shouldn’t really be a concern.
Humans are also a potential liability to a classified operation. If you can get the same results with 2 human analysts overseeing/supplementing the work of AI as you would with 2 human analysts overseeing/supplementing 5 junior people, it’s worth evaluating. You absolutely should never be blindly trusting an LLM for anything. They’re not intelligent. But they can be used as a tool by capable people to increase their effectiveness.
They use LLMs for what they can actually do, which is bullet point core concepts to a huge volume of information, parse a large volume of information for specific queries that may have needed a tech doing a bunch of variations of a bunch of keywords, before, etc. Provided you have humans overseeing the summaries, have the queries surface the actual full relevant documents, and fallback to a human for failed searches, it can potentially add a useful layer of value.
They’re probably also using it for propaganda shit because that’s a lot of what intelligence is. And various fake documents and web presences as part of cover identities could (again, with human oversight), probably allow you to produce a lot more volume to build them out.
That and an API (both things that there are search results implying they’re on the roadmap at some point) and my use would go way up I think.
It’s definitely not why I subscribed, but I’d definitely be all for options to use as many productivity tools as part of my subscription as possible.
It would be nice if it (at least optionally) was available separated into its own app at some point.
I honestly don’t know how. Any non Nintendo 3D game it’s instantly super obvious. And any game at all it’s obvious if you just try to navigate a menu with it.
They make the $10 Chinese junk alternates look amazing by comparison.
I’m not just talking about the drift. The actual intended product is ludicrously imprecise out of the box.
It changed gaming. The steam deck probably doesn’t happen without the switch. It clearly demonstrated the market for a handheld that manages large scale 3D games with complex mechanics.
But the joycons suck at more than connectivity. Straight out of the box those joysticks are fucking terrible, and they degrade hard and fast.
He explains that “cutting-edge AI capabilities” are now available for every company to buy for the price of standard software. But that instead of building a whole AI system, he says many firms are simply popping a chatbot interface on top of a non-AI product.
Well, yeah, because that’s what LLMs can do.
We’re not near the point where it’s reasonable or intelligent to allow “AI” into the driver’s seat. There are specific spaces where machine learning can be a useful tool to find patterns in data, and you would plug that model into normal tools. There are plenty of normal tools that can be made more user friendly with a well designed LLM based chatbot.
There are not a lot of spaces where you would want an ML model and and LLM interface, because there’s just too much extra uncertainty when you aren’t really sure what’s being asked and you aren’t really sure where the underlying patterns of the model come from. We’re not anywhere close to “intelligence”, and the people selling something claiming they’re “doing real AI” are almost certainly misrepresenting themselves as much as anyone else.
I’m always skeptical of anything LLM, but this looks like an interesting use case on the surface.
Fuck everything about blocking the hardware and hardware mods, but if it’s true that he was packaging games with it, you don’t even get to any of that, because they’re probably going to get every dime he has for “selling” their games.
I would much rather pay full price than still pay for a DRMed version that’s effectively guaranteed to be supporting some sort of organized crime group. Mass distribution at scale, with DRM, by definition means Russian organized crime, or a drug cartel, or some other global bad actor on that scale that’s doing shit like trafficking humans, arms dealing, drugs, etc, as well.
But ignoring that (and that I generally buy my content), I wouldn’t pay $.10 for an illegitimate copy that had an added layer of DRM on it. It’s fundamentally fucking repulsive for some subgroup whose whole business relies on bypassing someone else’s copy control to add their own.
DRM on pirated games is fucking gross as shit.
DNS names are restricted to your tailnet’s domain name (node-name.tailnet-name.ts.net)
I guess that’s fine for some. Not a compromise I’m willing to make though.
Because it’s a copy. It’s literally that simple.
Libraries can operate because of first sale doctrine. You can do almost whatever you want with a physical object that contains a copyrighted work.
What you can’t do is copy it. There is no possible legal way to distribute a digital copy of a work without an explicit license from the copyright holder. There isn’t even a legal concept of “owning” a digital copy. You purchase a license.
There’s really no credible argument that their distribution of books even might be legal.
Their only defense is fair use, and there’s no precedent for a “fair use” defense justifying copying a work wholesale for mass distribution. (Yes, “one copy at a time” to multiple people is mass distribution.) Copying a whole work has effectively only qualified as fair use when that copy is not re-distributed, and is actually for a personal backup.
The constitution explicitly grants authority to regulate IP. There’s absolutely no path to a constitutional issue, and constitutional issues are the only way you get laws overturned. “Other legal doctrine” means something like violations of due process somewhere in the chain, which is a constitutional issue, or direct conflict with another law.
The only possible judicial remedy is the premise that it’s fair use, which there’s a lot of precedent that it isn’t.
If it uses tags correctly you can just filter in and out what you want to see, then bunch by other common tags or whatever.
I have not reached the point of finding the right book hosting to properly self host my large collection of books, so I can’t really give a suggestion for a good browsing experience, but just generally speaking tags allow as much structure and organization as the front end wants to take advantage of. I’ve seen plenty of platforms that, once you pick your first tag, give a sorted list of other common tags you can dig down into, in addition to showing the list of content that meets the tag by whatever criteria you have. (An example I’m not sure exists, but very easily could, is to take the highest frequency set of tags with the least overlap (fiction/nonfiction/kids) and display them as titled shelves, then, once you click that, breaks down that group in the same manner until extra tags aren’t really useful.)
But in terms of the information they contain, the real world is fuzzy, so a method that allows for fuzzy buckets instead of strict ones is going to be more representative of the eventual content.