Jump to content

GluttonyReaper

Members
  • Posts

    525
  • Joined

  • Last visited

Reputation

497 Excellent

Profile Information

  • About me
    734

Recent Profile Visitors

6,519 profile views
  1. Right, this is pretty much the sticking point when it comes to all forms of machine intelligence - even in scientific fields, with purpose-built deep learning models for particular data sets, it seems that it's pretty tricky to cross that hurdle into finding new correlations that the average (or even expert) human wouldn't be able to find. It's great at reproducing some of the more nebulous correlations that humans can pick out, but catching the golden goose seems to require something different... ...which is currently the issue with AI-generated art. Theoretically it's only able to work within the parameter space given by its training set, and in practice it seems to only be able to work within a narrower 'average' portion of that parameter space where there's sufficient data to work with. I assume that's why it tends to produce either very 'safe' results that are arguably lacking that cohesive spark, or just weird nonsense that may as well be random, no matter what medium its working in. I had a friend actually who trained an older LLM (GPT-2?) on some old chat messages for fun before things really blew up... and results were a bit mixed. Whenever it was stuck with a small sample, it would either just regurgitate a message from the training data verbitim (or close enough) or spit out gibberish words that were only sometimes pronounceable. I was told at the time that this was probably due to a combination of the model actually generating character by character rather than word by word, and the fact that some of the words in the training data actually were made-up nonsense, because that's how people speak online sometimes.
  2. A picture would help here - orbits can bounce around a bit depending on craft design (e.g. minor flexing between parts can throw off the calculations a little), but it usually evens out when you let things rest. Does the orbit stabilize when you timewarp?
  3. I actually had the opposite issue on my last machine - no matter what resolution I set it to (usually 720p as I was trying to boost performance), it would inevitably reset to my native screen resolution (3000x2000). It never switched from fullscreen to windowed though, that's a new one...
  4. For reference, this would put Mars viewed from Deimos at about 33 times wider than the Moon on our sky. For Phobos, it's almost 80 times wider. The other thing that occurs is that even during the 'night' on either moon, if you were on the side facing Mars you'd have this pretty large thing in the sky reflecting a fair amount of light. So weirdly enough, presumably the darkest time would actually be in the middle of the 'day' during your monthly eclipse.
  5. Obviously I can't comment on the quality of Science, but this is pretty the much main thing that's keeping me from picking up KSP2 - sure it looks like it's technically a better game than KSP1, but there's yet to be anything that really sticks out a "standout" feature that actually pushes me over the line. Streamlined systems and quality-of-life stuff like procedural wings and thrust-on-rails (while presumably a lot of work, and I'm happy they exist!) are obviously improvements, but they're not exactly enough to convince me spend £45 on the game when I already own KSP1. The only thing which really sticks out is the (vastly) improved graphics which, while essential for a game like this, can't really carry the whole thing. (I should also clarify that buying KSP2 would involve buying a new machine that can actually run it... so I may be a little biased in that regard) In general it feels like what we've seen so far isn't anything that couldn't have been implemented in KSP1 if it had continued development (game design wise at least), and any of the interesting features like colonies, resources, interstellar, etc. are so abstractly described that it's difficult to get excited for them. I had really hoped that a fresh start with a more professional team would have lead to all the stuff great about KSP1 being distilled down, and everything else being replaced entirely with something more game-like. Instead it feels like there's been a lot of time spent faithfully recreating KSP1, warts and all... just to continue where the first game left off, running into a lot of similar issues unsurprisingly. That said, I would also note that there doesn't seem to be much pressure from the KSP2 team to actually sell the game at this point. Other than the initial EA push (which seemed to be a bit of mess), hype and marketing still seems to be contained to existing community rather than trying to bring loads of new players in. Perhaps they're aware of how bare-bones things are at the moment, and are waiting until there's more to show closer to a 1.0 release before really pushing it out there.
  6. Somewhat tangential to the topic of extra planets (although personally my stance is that I'd rather have outer planets instead of interstellar travel...) but I'd quite like to see the asteroid system expanded to include a lot of these minor bodies, seeing as many of them are usually on the order of only a few kilometres in size. It'd be pretty neat to have to find and track the minor moons in the same way as we do with asteroids in KSP1, and that only leaves a handful of hand-made bodies to make for each planet, which lets you keep them nice and diverse.
  7. They probably should, yeah - but I still feel like the convenience of them is enough to make them worthwhile without hobbling other engines, in a way that makes the game a fair bit more complex. I disagree on the RCS front... I'm a big fan of the game letting you make your own mistakes. Mainsails are obviously a terrible choice for RCS control, but that's something the player should find out themselves in a fun way, rather than the game arbitrarily deciding which engines "count". Plus, there's no actual reason that someone couldn't use Terrier engines as RCS thrusters on a really large craft, and that's definitely not something I want the game to stop people from doing.
  8. Having messed with mods in the past that add similar mechanics... I'm kinda split on this. I like that it makes big lifter engines feel a lot more powerful when you have to let them spool up to full throttle before launching, but it does get a bit old after a while if you're launching a bunch of rockets, and I'm not really sure it fits the stock "vibe". Sure, but arguably KSP1 managed to let most engines exist pretty comfortably in their niches without this quite successfully - I'm not really sure it's worth restricting them even further just to justify the existence of manoeuvring engines. In general, I think that KSP should err on the side of generosity rather than tight restrictions, and this definitely feels it would start to walk down the path of "There is one objectively correct way to build you spacecraft, and any deviation will be heavily penalised". That said, I wouldn't be opposed to there being a minimum thrust limit on some engines at least... it was always a bit odd to me that KSP1's infinite thrust limiting meant you could theoretically use your lifting Mainsail for docking. No opposition to this, though, if anything the game should let you tie any engine you want to RCS. The opposite would be great too, for smaller craft there's no real reason you couldn't tie your RCS engines to your main throttle instead of having any liquid fuel.
  9. And even better - if you're like me, and not really a fan of far-future interstellar tech in KSP, you can just... not use it. The devs so far seem pretty committed to creating a game where the player isn't forced to do anything arbitrarily.
  10. Right, at the very least, you presumably don't want to use up your rare precious resource when collecting more of that resource...
  11. Oh sure, mass extinction events could act as a "genetic reroll", giving a chance for creatures to evolve that never would otherwise... but there's no guarantee that they'll be better (or worse) in terms of, intelligence for example, than their predecessors. As always, it's difficult to gauge whether it ends up being a net positive versus letting life evolve through the slow route, because our sample size for development of intelligent life is still just 1 (for a particular definition of intelligence)
  12. Right, it's closer to dousing your Petri dish in antibiotics than most natural causes. It's likely to be more of a man-made mass extinction event than anything else. Depends what you mean by "greater" Arguably, the surviving species would simply be differently adapted rather than better in an absolute sense - they might be better at surviving in a more variable climate, but at what genetic cost? In general, I assume the massive drop in biodiversity would outweigh any gains made by individual remaining species - e.g. previous mass extinction events mean we don't have any giant lizards or insects running around anymore (although that might have been inevitable anyway).
  13. What we really need is an Austin-Powers-esque cutscene that plays every time a Kerbal dies, in which a space program ambassador has to explain to their spouse and children how they were killed because someone forgot a parachute again... /j
×
×
  • Create New...