• 2 Posts
  • 58 Comments
Joined 1 year ago
cake
Cake day: July 21st, 2023

help-circle
  • Ttereal tellers is ttattElonkows nothing about AI. Anyone involved in the field knows all of the big names because we read their papers, listen to their lectures, and talk about their models. He then goes on to be dismissive of work he’s not even close to understanding. It’s blatant ignorance, and Elon is used to just being able to power through his ignorance by either BSing his way past people who know no more than him or firing anyone who is actually qualified and as a result disagrees with him.


  • This the order in which you should try to access papers:

    1. Normal Internet search including quotes to force the title and components like “pdf”
    2. Organizational/lab pages of the authors. Very many people will put either full papers or preprints on their personal professional pages.
    3. Preprint services like arXiv. The ones you look at will be determined by subject area. Preprints will usually only differ from the published work in formatting.
    4. Just email the authors. Most of us are so happy that virtually anyone wants to read the paper we spent months on that we will happily send a copy. Because people are busy you might need to hit them up a couple of times, but most will be more than happy to send you a copy, and most publications specifically carve out to allow authors to do that.




  • I’ll change my socks every day, and more if I’m swapping out workout clothes for fresh clothes.

    Oddly, I really don’t experience foot odor. Other bits can get rather ripe, but whether I’m wearing boots or going barefoot, my feet just do t get funky. That said, I have had athlete’s foot and I’ve seen what happens when someone doesn’t change their socks after days of marching and working in the field. I know it’s not a major danger for me at this point, but I’ve got a whole drawer full of socks, and in any case I want them to match with my shoes and pants.



  • That’s also what general happens in the US if you die without a will. It’s called intestate succession. If your spouse survives you, they will generally get everything held as community property. If you had property prior to the marriage, it might be divided among your children, if any. Of there are no immediate family (spouse and/or children), it can then be divided among any relatives you do happen to have.

    If you have any valuable assets that you want to bequeath to someone in particular, then it’s important to have a will. If you want a bigger chunk of money to go to a relative you know is struggling financially or could otherwise use it (kids heading to college, buying a house, whatever), then that’s another reason to have a will. Of instead of a family member, you want the money to go to a charity, then that’s yet another reason.


  • I bought the steam deck with the largest storage available, and still ended up picking up a 1Tb SD card. I wouldn’t end up tossing that idea unless you have a modest library without a lot of the big footprint games.

    I’ve reached that point where I’ll spend some extra money to avoid the inconvenience of having to redownload a game if it’s at all unavoidable, and the internal storage gets filled up a lot faster than you’d think. You know your own library and habits, but you might be happy to have one on hand.




  • I’m not even sure what he’s talking about. Open access journals are the ones who charge authors to publish.

    If you publish in a journal that has closed access, there is generally no fee to publish. If you want your paper to be open access, you can tack on an additional open access fee so that your paper doesn’t end up behind a paywall. The last time I looked - and this was several years ago - the going rate for making your paper open access in a closed access journal was about $2-3k. We always budgeted for publication fees when we were putting together our funding proposals.

    The fee structure is similar for open access journals, except that there’s not a choice about paying them. For researchers whose work isn’t grant funded, it generally means they’re paying out of pocket, unless their institution steps in.

    I had a paper published in a small but (in its field) prestigious journal, and the editor explained to me that he only charges people who can afford it, and uses those funds to cover the costs of the journal. He explained that he had a paper from a researcher who couldn’t cover the publishing fee, and he let me know that I was helping out the other person, too.

    What I don’t understand is how anyone how has gone through academia doesn’t know this.






  • The oldest crpg I ever played was called advent, because the Vax computers could only use 6 characters for file names and so the people who ported it couldn’t use the actual name “adventure.” It was basically the same as the game infocom shipped as Zork.

    Apparently the original implementation was on the PDP-10 in 1976. There might have been a couple other games that predated it by a year or two, but adventure was the big one in my opinion because it led (eventually) to the creation of the infocom text based game engine and a whole line of games ranging from hitchhiker’s guide to the galaxy to leather goddesses of Phobos.


  • I do not mean this to come off as blunt as it sounds, but I’m trying to be both clear and concise.

    What you’re talking about is not how game theory works. What you’re doing is taking the most basic, highly abstracted representation of a generic idea and expecting it to correlate with reality. It’s the same thing people do when they ascribe some kind of wish fulfillment to the free market or to evolutionary dynamics. It’s not even a platonic ideal - it’s drawing a supply/demand curve and thinking you understand how prices work in a market economy. Here’s the main issues you’re running into when you try to play Ultimatum with something the size of the Democratic Party:

    1. Noise. There is a permanent base of 3-5% of the electorate that’s going to vote Green, or whatever. The protest voters almost never rise above that noise floor. Focus on a single (potentially complex) issue would help. Green rallies (and others) often have everything from antivax to prison reform to the environment to voting rights to BDS and BLM. All of those things (except the antivax) might be important, but there needs to be a central focus. IMO it’s voting rights - I’d love DSA to drop everything to just start suing states and protesting for voting rights, because everything else is lost without that. We can even both/and, as long as there’s a vision and a focus on a main first objective. Right now we’re coming off like a bunch of verses from We Didn’t Start the Fire. Ultimatum with multiplayer and a noise function is a completely different game.
    2. Feedback loop. The consequences for actions needs to be tightened up, and they need a wide base. There needs to be visible and constant representation out in front of both cameras and politicians. This can be people like the Squad or figures like Robert Reich, but there needs to be a uniform voice that doesn’t wait for the election cycle. Groups like Moms for Liberty have this kind of thing on lock. They have a brand and spokespersons and will host and endorse, or else attack on Fox News within hours of a political decision. They’re shit in every way, but they can work the machine. Ultimatum with a delayed feedback loop is a completely different game because the failure of the deal is less attributable.
    3. Solidarity and messaging. The majority of Americans want universal health care. The majority of Americans want green energy. The majority of Americans want a cease fire in Gaza. By spreading opinions across multiple realizations of this top level policy objectives, we dilute the message. Ultimatum requires identifiable players with identifiable agendas.

    We as voters aren’t playing Ultimatum. Instead, we are playing minimax as an emergent strategy to defend the rights of marginalized populations.


  • I am a moderately heavy kindle user and have been since the second version they shipped. When I upgrade, I usually buy the best new model available. I am skipping the one with pen support because Amazon’s text autosuggestions are absolutely the worst I have ever seen - it’s like they’re just using a random number generator and not a predictive algorithm - so my current Kindle is the Oasis.

    It is so far beyond any other one I’ve owned that they’re not really comparable. The backlight is steady and even with no patchiness. The text reads cleanly with no fuzziness around the fonts. It’s comfortable to hold, and because it just inverts very cleanly and automatically it makes it trivial to hold upside down if you change hands or roll over. My requirements for a case are that it makes the device easier to hold and prop up for hands free reading in bed. Any of the origami cases should do - I think they’re all very similar in design but I’d just go off the reviews for build quality.

    That said, there’s a number of kindle books that cannot be read on kindle devices because the publisher decided to prioritize the formatting over the text, and those I have to read on one of my iPads. I still prefer the kindle for text only books because it’s lighter and easier to hold.

    The oasis has a slightly different form factor so it might be worth checking out in person, but I went from skeptical to really appreciating the design.


  • That is literally not how it works. That’s how people think it should work, but when you see that it doesn’t, you have to turn back and review your premises and your model. I know the way you think it should work and how you want it to work, but when it doesn’t work you need to revise.

    The problem is this - the feedback loop is insufficient and the correlation is unclear. If you are directly negotiating with someone, then you can play Ultimatum. If you are one of a hundred million people casting a vote for one person or another, you cannot. Perot cost Bush I the election, and Nader cost Kerry the election. Neither party decided that they needed to move in the direction of the spoiler candidate. They’re especially not going to do so for 3p candidates who pull in the low single digits, even if they lose by low single digits, because they’ll think they can get more by moving towards the center.

    You can vote however you want, but don’t base it on a theoretical foundation that has less than zero application to the scenario you’re modeling. It really, honestly is a minimax choice, and if you are truly an ally for those of us in marginalized communities, you have to recognize it.

    I’m not being a right winger here - I’m a member of the DSA and this is in line with what they (and people like Chomsky) advise. But I’m not talking about even that angle. I’m just talking minimax and BATNA. If negotiations fail (ie we didn’t get Bernie), the best alternative is Hillary. At least Roe wouldn’t have been overturned and we wouldn’t have states suing to make ten year olds give birth to their rapist’s babies.