• 0 Posts
  • 30 Comments
Joined 2 years ago
cake
Cake day: July 8th, 2023

help-circle
  • I made the attempt, but couldn’t parse that first link.

    Fair - it is indeed difficult for non-experts. But all you need to see from it is that it is a concrete example of a (small) actual quantum computer as reported on outside a corporate press release. The focus on error correction comes from the fact that this is the next big hurdle in the way of scaling up. But the machine is there!

    This is just one more kind of chip that will be found in computers of the future.

    Exactly - this was never meant to replace classical computers, but to do things that are impossible for classical computers to ever do.

    Problem is, this only works for systems that have a known answer (like cryptography) with a verifiable result, otherwise the system never knows when the equation is “complete”.

    This isn’t quite right. It’s true, there’s never 100% certainty you have the right answer, but 99.99999% is usually good enough. A classical computer also isn’t 100% certain since it’s also technically just a “physics experiment”, but it has an extremely low error rate, like 10^(-20).

    when they talk about speed, they aren’t exactly being forthright

    Sure, quantum computers aren’t faster than a classical computer for now, and won’t be for a while. But exponential speedup means that the problems we can eventually solve with a quantum computer are literally impossible for a physical computer to ever solve. This part of the corporate hype speak is true. It’s a purely physical fact. Though for sure we aren’t there yet!

    it’s… not really useful in power expenditure or financially to do much beyond a large corporation or government breaking encryption.

    Indeed, very likely nobody is ever going to be doing personal computing on these, but they were never meant for that, they are meant for supercomputing level calculations.



  • Hell yes! I’d love to share some stuff.

    One good example of a quantum computer is the Lukin group neutral atoms work. As the paper discusses, they managed to perform error correction procedures making 48 actual logical qubits and performing operations on them. Still not all that practically useful, but it exists, and is extremely impressive from a physics experiment viewpoint.

    There are also plenty of meaningful reports on non-emulated machines from the corporate world. From the big players examples include the Willow chip from Google and Heron from IBM being actual real quantum devices doing actual (albeit basic) operations. Furthermore there are a plethora of smaller companies like OQC and Pasqal with real machines.

    On applications, this review is both extensive and sober, outlining the known applications with speedups, costs and drawbacks. Among the most exciting are Fermi-Hubbard model dynamics (condensed matter stuff), which is predicted to have exponential speedup with relatively few resources. These all depend on a relatively narrow selection of tricks, though. Among interesting efforts to fundamentally expand what tricks are available is this work from the Babbush group.

    Let me know if that’s not what you were looking for.












  • Well, if by AI you mean large language models, they tend to do better at language tasks than math tasks. So a better example might be that it’s easier to get an LLM to write a statement for you and checking if it’s correct than writing the statement from the bottom.

    The square root was just a clearer example. In the case of OP, it might very well be easier to have an LLM propose relevant case law and then check if that case law exists and is relevant, rather than having to find it yourself from square one.





  • Giving it a replay on realistic mode, it’s good fun! I like not fast traveling, and having to figure out where you are by actually reading the map is fun.

    It has its jank, and the silly macho-vibes suck, but it’s a really interesting and unique game. Try to lean into its oddities, they actually work quite well when you give them a chance. The saviour schnapps saving system is not that punishing, but it does mean you either have to commit your early economy to saving/stealing schnapps or accept that you will fail quests and die in combat without being able to quickly reload a recent save.

    Ultimately, I enjoy games that force me to take failure seriously, but that’s a matter of taste. It does make riding through the woods pretty intense with the threat of getting ambushed.


  • Try to use the combat skills you’ve unlocked instead of spamming thrust.

    General combat advice:

    • Train your battle skills with Bernard(?) at the training ground. Both to get some XP and to learn some important moves other than spamming thrust.
    • Master stroke (?) is a really useful riposte move which will make things a lot easier.
    • In general, avoid taking on more than one opponent until you’re well-armored
    • Use the right weapon, ideally a type you’ve both trained with and which is effective against your type of foe, e.g. warhammer for armoured foes.

  • I’ve bought each of the three latest Hitman titles on Steam exactly once at regular video game price and gotten all the content of each game, plus freely received the collection of all that content in the “World of Assassination” consolidation. Plus the free and amazing rogue-like expansion taking place across all maps from all games.

    Seems like a really good deal to me.