cdb's recent activity
-
Comment on What words do you recommend? in ~talk
-
Comment on Why do some gamers invert their controls? Scientists now have answers, but they’re not what you think. in ~games
cdb Link ParentAlthough ADHD isn't specifically one of the factors in the paper, it seems like they did ask some questions in the survey that are similar to questions on ADHD questionnaires. This was represented...Although ADHD isn't specifically one of the factors in the paper, it seems like they did ask some questions in the survey that are similar to questions on ADHD questionnaires. This was represented by the "ITQFocus" item, which didn't predict invertedness preference.
Some question examples out of about 30 along similar lines:
Do you ever get extremely involved in projects that are assigned to you by your boss or your instructor, to the exclusion of other tasks?
How easily can you switch your attention from the task in which you are currently involved to a new task?
-
Comment on Why do some gamers invert their controls? Scientists now have answers, but they’re not what you think. in ~games
cdb (edited )Link ParentI'm pretty sure that whether you invert or not has nothing to do with left or right handedness, and it was just an off the cuff example to say that people may not know if they're using the type of...I'm pretty sure that whether you invert or not has nothing to do with left or right handedness, and it was just an off the cuff example to say that people may not know if they're using the type of controls that best suits them until they try it out both ways. The research article doesn't seem to mention handedness as a factor in invertedness preference, other than that they recorded it.
Edit: my crtl-f failed me. It does have handedness in the results, and it's not a strong predictor of inversion preference
-
Comment on I made a tool to generate AI powered recaps of TTRPG sessions in ~games.tabletop
cdb Link ParentI know OP has already answered, but I've tried a similar thing with live sessions by recording audio with my phone, using the on device transcription, then giving it to Gemini, prompting to...I know OP has already answered, but I've tried a similar thing with live sessions by recording audio with my phone, using the on device transcription, then giving it to Gemini, prompting to generate summaries with the context of a DND session, keeping in mind that it's a transcription with errors, and asking it to ignore irrelevant banter. The summaries were really good, including a comprehensive rundown of what happened in the campaign while ignoring all the side chatter. I didn't ask any specific questions about the session though.
-
Comment on Lung cancer rates rising among non-smoking women in ~health
cdb LinkToo bad we don't have a good idea why this might be. All the possibilities suggested in the article are just speculation. Hoping for some better data soon. I don't know if they deliberately...Too bad we don't have a good idea why this might be. All the possibilities suggested in the article are just speculation. Hoping for some better data soon.
I don't know if they deliberately focused on these two people to have some balance of severity of the problem and optimism, but I found it interesting that both are in bad situations that were probably death sentences not too long ago but are expected to live in the near term. It's kind of amazing that we have effective treatments for these kind of things nowadays, although we wish they were better.
-
Comment on Lung cancer rates rising among non-smoking women in ~health
cdb (edited )Link ParentIt's a frustrating situation, but given that the fetus is very often considered the greater priority by the "incubator" themselves, I'm not sure how we can solve this problem until we get better...It's a frustrating situation, but given that the fetus is very often considered the greater priority by the "incubator" themselves, I'm not sure how we can solve this problem until we get better
ex-vivomodels for this kind of thing.Mostly just addressing the idea that society and drug developers are ignoring this important issue. It's a topic that gets a lot of attention during drug development. It's more that we haven't thought of a moral way to solve the problem.
Edit: realized that specifying ex-vivo could imply use of fetal tissue, which I didn't mean to do. Could be any models, really, as long as they're good.
-
Comment on Lung cancer rates rising among non-smoking women in ~health
cdb Link ParentIt's a really hard problem. There's no good way to really ensure things are safe for expecting mothers except dose a bunch of them and see if anything happens to the fetus, which is often not a...It's a really hard problem. There's no good way to really ensure things are safe for expecting mothers except dose a bunch of them and see if anything happens to the fetus, which is often not a road people want to go down for obvious reasons. There are tests that are performed to flag teratogenicity, but the body is a complex system that we don't fully understand. In terms on morality, the potential harm needs to be balanced against the potential good, so if there's any indication whatsoever of possible teratogenicity you might still feel like you need to exclude certain populations to help others. I guess my point is that although it is a kind of sexism, it's wholly due to practicality rather than malice or profit-maximizing. Drug companies would really prefer not to have to exclude significant populations on their drug labels. It's unfair that women have to work harder to make babies or just deal with the capability to make babies in the first place (periods and such), so life is sexist to begin with.
Pain is an interesting example. It's hard to test acute pain in the clinic without using elective procedures, so the trials for a new pain drug approved earlier this year used abdominoplasties (tummy tucks) and bunionectomies. This meant that the study population was almost all women, so it's men that we don't understand the effects on for this drug. Since the drug is being prescribed to the general public, I guess we'll find out soon.
-
Comment on Does anyone here enjoy a whisk(e)y ? in ~food
cdb LinkI like whisky enough that I went on a distillery tour of Islay this year. We visited Bunnahabhain, Bruichladdich, Kilchoman, Bowmore, Laphroaig, and Lagavulin (Ardbeg was closed). Bruichladdich...I like whisky enough that I went on a distillery tour of Islay this year. We visited Bunnahabhain, Bruichladdich, Kilchoman, Bowmore, Laphroaig, and Lagavulin (Ardbeg was closed). Bruichladdich was my favorite. I'm a fan of heavily peated whiskies, so the single cask Octomore I tasted was my favorite. I'd love to drink more of the stuff, as they have ~3 versions each year, but at $200-300 a bottle it's a pretty big investment in something with steeply diminishing returns on price like whisky. The Port Charlotte 10 and Ardbeg Uigeadail are still pricey but slightly more affordable alternatives that I can have more often. Another whisky I liked recently was Benromach 10, which is a lightly peated speyside whisky aged in bourbon and sherry barrels.
-
Comment on Does anyone here enjoy a whisk(e)y ? in ~food
cdb Link ParentI enjoy Redbreast pretty often. It's nice when Costco has it in stock. I tasted the Redbreast 12 and 15 side by side and actually preferred the 12. I think the 15 is more refined and balanced, but...I enjoy Redbreast pretty often. It's nice when Costco has it in stock.
I tasted the Redbreast 12 and 15 side by side and actually preferred the 12. I think the 15 is more refined and balanced, but also somehow more boring. I guess it saves me some money since I don't feel the need to go for the more expensive bottle.
-
Comment on Sliding scale payment systems in ~talk
cdb LinkSeems like kind of a half-measure that might cause all sorts of headaches with difficulties in setting prices, task priorities, client selection, and all the other stuff people have already...Seems like kind of a half-measure that might cause all sorts of headaches with difficulties in setting prices, task priorities, client selection, and all the other stuff people have already mentioned. I'd prefer to get paid what I'm worth for my work, then donate money or volunteer for organizations that are doing some good locally. Trying to fit somewhere in between sounds like more trouble than it's worth.
-
Comment on Is the concept of debate completely useless? in ~talk
cdb (edited )LinkHave you ever changed your mind when debating with someone? Or have you at least acknowledged that the other person had a good point or good bit of information that might alter your opinion or...Have you ever changed your mind when debating with someone? Or have you at least acknowledged that the other person had a good point or good bit of information that might alter your opinion or make you walk back some portion of your argument? If not, maybe the problem is how you approach the debate.
I know there's a bit of selection bias where people are more likely to join in if they already think they're right, so I think it makes sense that most of the time you wouldn't think that your opinion is changed by a given internet argument. However, if one doesn't go in with an adversarial attitude that "I joined this debate because I'm confident I'm right, so everyone who disagrees with me is stupid," it should occasionally happen that the other person makes a good point, even if it doesn't completely flip your view. I think that much is the best we can hope for from casual debates.
As for the other person, I've given up on thinking I'll ever get anyone to say "wow, I never thought of it that way, I was wrong and you're right," and I think it's a good thing not to approach discussions that way.
P.S. What a great topic. So many people are dying to weigh in here. It's almost like group therapy. Maybe it's because we can't help but argue that all the time we've spent typing comments to strangers on the internet has been worthwhile. Otherwise....
P.P.S. Everyone who disagrees with what I've written here is wrong.
-
Comment on 2025 NFL Season 🏈 Weekly Discussion Thread – Week 1 in ~sports.american_football
cdb LinkI usually watch redzone, but I didn't feel as entertained by it last Sunday. I don't think it's the ads, more that you don't really get the context or narrative behind anything. I might switch to...I usually watch redzone, but I didn't feel as entertained by it last Sunday. I don't think it's the ads, more that you don't really get the context or narrative behind anything. I might switch to actually just watching a game instead.
I'm a little sad that I missed the end of a really good game between the ravens and bills. I turned the game off and went out to dinner after it looked like the ravens were winning. Apparently it was a really exciting 4th quarter.
-
Comment on 2025 NFL Season 🏈 Weekly Discussion Thread – Week 1 in ~sports.american_football
cdb Link ParentI think giving up on the dolphins after just one bad game is an overreaction. Bad games happen. At least you didnt give up 70 points like when you destroyed the Broncos 2 years ago. Last season...I think giving up on the dolphins after just one bad game is an overreaction. Bad games happen. At least you didnt give up 70 points like when you destroyed the Broncos 2 years ago. Last season was barely a losing season at 8-9 after four straight winning seasons. Tua has seemed pretty good when he wasn't injured for the past few years. If he can avoid another concussion, you have a relatively easy schedule based on Vegas odds, so you should still have a decent shot at making the playoffs. Week 1 can be a little crazy. I'd say wait a few games before writing them off.
I'm just a Raider's fan, though. I'm used to losing.
-
Comment on Don’t like joining in? Why it could be your superpower. in ~life
cdb LinkI felt that this article was really light on details about what makes this a "superpower". Sure, self-acceptance and independence are good, but far from "super". If anything, we have the opposite...I felt that this article was really light on details about what makes this a "superpower". Sure, self-acceptance and independence are good, but far from "super".
If anything, we have the opposite problem. The article says:
Writers, thinkers, policymakers, and even the US surgeon general have cited the decline of communal life as a principal cause of poor mental health and proposed solutions ranging from getting off social media to expanding our social support networks. In theory, these ideas are not without merit. But in practice, we are having more and more conversations about the importance of community, while continuing to become lonelier and more divided than ever.
It's not that encouraging socializing is counterproductive. It's starting to look like more of a problem, so it's getting more attention over time. To me, this just means that people are getting better and better at withdrawing from society, and I'm not sure the best move is to encourage it.
While I'm happy to see people being more in tune and honest with their feelings these days, it's also good to realize that your feelings are not always the most important thing. To me, maturity is realizing that sometimes your feelings should be secondary in importance while you do what is practical and beneficial for yourself and others. If you never accept invitations from friends, you'll eventually stop having friends. Some people feel ok with this, but if this is a widespread phenomenon, it might not be the healthiest thing in the long run for the individual and society overall. If it's the norm to isolate yourself, then you can't even be an occasional participant to a friend group, because there aren't enough friends groups anymore.
I myself am an introvert, and looking back at my life, I see detrimental effects from excessive social isolation at certain times in my past. It took a lot of work and therapy to get to a better state (not extremely social, but not extremely isolated either). I'm not saying that I'd become a social butterfly if I had a chance to do it all over again, but saying "yes" to invites just a few extra times would have been a huge improvement for sure. Relating this back to the article, if you'd asked me at 24 what my opinion was about my social life, I would have said that it was fine. We just don't realize how much social interaction happens automatically at school and jobs in college towns. There's a bit of a residual where these friendships can carry on somewhat automatically for a few more years, then it takes effort to keep them going. Years later I looked back and realized that my deprioritization of socializing after graduation was a bad decision.
-
Comment on What's your go-to hot sauce? in ~food
cdb (edited )LinkI bought a variety pack of Melinda's hot sauce, and I've enjoyed the ones between 2-4 out of 5 in spiciness rating. The 1 is too mild to be called hot sauce and the 5 is just excessive for my...I bought a variety pack of Melinda's hot sauce, and I've enjoyed the ones between 2-4 out of 5 in spiciness rating. The 1 is too mild to be called hot sauce and the 5 is just excessive for my taste, more suitable for cooking rather than seasoning your individual portion.
Another one I use pretty often is a chili crisp that my dad makes using real Sichuan peppercorns and dried chilies. I guess that's not that helpful for you, but there are some decent ones available at Asian grocery stores. Or if you're feeling frisky you can buy the ingredients and make your own.
I don't use a ton of hot sauce though. I tend to just add spiciness while I'm cooking, but I guess that doesn't work if not everyone eats spicy food.
-
Comment on Why do you like your job? in ~life
cdb Link ParentOn a slight tangent, although I really like working from home sometimes, I think less social chatter is probably the biggest problem with remote work. We're reducing our forced social interactions...On a slight tangent, although I really like working from home sometimes, I think less social chatter is probably the biggest problem with remote work. We're reducing our forced social interactions on all fronts, with online everything. This results in lower social skills, fewer friendships, and just less social interaction overall. In the end, we get worse mental and physical health outcomes and other issues like polarization of opinions (you might not think of people who disagree with you as real people unless you have a conversation with them in person).
I'm not saying it's great to spend all your life at work, but the incidental interactions of being there in person definitely has a significant function in society that isn't being replaced.
-
Comment on Tildes Book Club discussion - August 2025 - Cats Cradle by Kurt Vonnegut in ~books
cdb Link ParentThis was my first Vonnegut book. It was the biggest load of nonsense I've ever read. It was also one of the most fun and thought-provoking things I've read. Every time a new character was...- Exemplary
This was my first Vonnegut book.
It was the biggest load of nonsense I've ever read.
It was also one of the most fun and thought-provoking things I've read.
Every time a new character was introduced, I'd feel like it was just more randomness/quirkiness for the sake of being odd, but after putting the book down I wouldnt be able to stop thinking about it. I'm not sure if I'm dumber or wiser for having read it, but I'm glad I did.
-
Comment on I asked Michelin chefs how they cook ramen in ~food
cdb Link ParentOther than the gnocchi one, I felt like you could just do the easiest part of what they did and have a much better result than the typical prep method. It seemed like sautéing onions with the...Other than the gnocchi one, I felt like you could just do the easiest part of what they did and have a much better result than the typical prep method. It seemed like sautéing onions with the flavoring packet was a popular strategy. For the third one, I would have just stopped after adding the noodles. It was a reminder how quick and easy something like a kimchi jjigae could be. You just need kimchi, scallions, pork belly, frozen tofu, and a ramen packet. I also felt like the last one was very realistic, assuming you sub in bacon or mushrooms instead of A5 wagyu brisket.
-
Comment on If you're a programmer, are you ever going to believe an AGI is actually 'I'? in ~tech
cdb Link ParentI'd take what you're saying a step further and say that I don't think that even having the skillset of programming AI models will help one evaluate whether something is "intelligent" or not....I'd take what you're saying a step further and say that I don't think that even having the skillset of programming AI models will help one evaluate whether something is "intelligent" or not. That's more of a philosophical argument. I'm not working on the latest models, but I can calculate gradient descent by hand and create BERT or similar in pytorch, while understanding what each part is doing. None of that helps me evaluate whether a model is intelligent or whether we have reached "AGI."
-
Comment on Question - how would you best explain how an LLM functions to someone who has never taken a statistics class? in ~tech
cdb LinkI feel like this touches on what I think is a really hard problem, which is that randomness touches almost every aspect of our lives, but humans don't find randomness intuitive. People will think...I feel like this touches on what I think is a really hard problem, which is that randomness touches almost every aspect of our lives, but humans don't find randomness intuitive. People will think that a 95% chance is basically guaranteed and cry foul if the 5% outcome happens. Even people have taken a statistics class and can draw out the probability table for the Monty Hall Problem may argue when given the correct solution. I have a decent amount of education and experience with randomness and probability, but I feel like I still don't truly have good intuition about it. I've just learned enough to distrust my intuition and trust the calculations. It's pretty interesting that I can still go on tilt and get emotional when playing poker when losing hands I "should have" won, despite knowing that low probability occurrences are inevitable.
When you talk about the worst problems of LLMs, I assume you're talking about the stochasticity of their outputs. As far as outputs, I'd compare an LLM to a person who has read everything on the internet. They're computer programs that have learned a bunch of things whether correct or incorrect, and they output words that are often correct but sometimes incorrect.
I'm sure everyone can think of a time when they or someone they know said the wrong thing. Maybe you know better, but people are not perfect and sometimes you have a brain fart and say the wrong word or explain some concept incorrectly. Another example might be how sometimes a friend is telling a story, and your memory of the story is a little different. It's not possible for both of you to be right, and it's actually possible for both of you to be wrong. Maybe one or both of you mixed up or mashed up that memory with another similar memory, so the retelling comes out wrong. There are also times when maybe you learned about something in 3rd grade, only to find out that this was completely false later in life. That doesn't mean that your 3rd grade teacher was a terrible person, just that people are fallible and sometimes learn and propagate incorrect information.
Similar kinds of things can happen with LLMs too. Although they're often really good at answering questions, they're sometimes wrong, and their outputs shouldn't be counted on as authoritative sources. We shouldn't trust any given human to be an authoritative source either. An expert human is only usually right on their topic of focus. If an accurate answer is needed, it should always be backed up by some kind of authoritative information.
I know people never agree on these things since they can be regional, but as an American, I'd like to suggest "POR-tuh co-SHARE".
You can't just say "port" because us lazy Americans end words like this with a glottal stop rather than actually pronouncing the "t" at the end. Also, "chere" in French sounds more like "share" in English. Finally, emphasis was added using upper case to de-emphasize the "tuh" and because words like "co-chair" or "copilot" always have emphasis on the "co" which would be wrong here.