curvise

Karma: 44

curvise 17 Jan 2026 6:04 UTC
1 point
0
on: How AI Is Learning to Think in Secret
This could be obvious to most people here but can you briefly explain how Neuralese is a new thing and not just “how LLMs worked before COT was invented”?
Even without understanding that though, I found this post excellent at catching me up on the topic!

curvise 6 Jan 2026 6:28 UTC
4 points
1
in reply to: RationalElf’s comment on: In My Misanthropy Era
I think this is an important point a bit buried here. I like my therapist and listen to her and want her advice—she’s not just a sounding board. However, I am pretty sure I’m smarter than her, and I’m definitely more oriented toward truth-seeking, which doesn’t seem to be something she really thinks about much. And yes, I do find myself frustrated and contemptuous during some conversations. But I continue to see her because I trust that she still has a lot to teach me! I’d be a terrible truth seeker if I saw her fumble one conversation and decided that she couldn’t help me in any way.
So I guess the important thing is that I don’t engage her in philosophical debates? Jenn, perhaps the answer is to meet with “common” folks on their own ground. It strikes me while writing this that the philosophy meetup you went to might have been the worst possible setting to test your empathy. These people were racing on hands and knees to call themself the fastest in the world and had never even heard of running.

curvise 29 Dec 2025 14:59 UTC
4 points
0
in reply to: bfinn’s comment on: Arithmetic is an underrated world-modeling technology
if you’re checking how many fruits you have, “apples and pears” makes sense.
If you’re trading a bag of apples for a bag of pears, you might want to know the relative value of apples and pears, so you would indeed calculate “apples per pear”.

curvise 15 Nov 2025 9:32 UTC
6 points
14
in reply to: Steveot’s comment on: Everyone has a plan until they get lied to the face
I also thought it sounded… really annoying. something I may have found interesting 10 years ago but would now cause me to simply avoid the person. And it might ruin my night by making me feel like a party pooper, a la Thane’s comment above.

curvise 13 Nov 2025 16:35 UTC
1 point
0
in reply to: Richard_Kennaway’s comment on: A glimpse of the other side
Those are good points. I was expecting something different from this post but only based on my intuitions, not explicit framing.

curvise 13 Nov 2025 5:32 UTC
1 point
0
in reply to: Richard_Kennaway’s comment on: A glimpse of the other side
I suspect you’re way off the mark here. I downvoted this post because it felt like magical thinking. “Just break your phone addiction and your life will be exactly what you want” is not true. But that seems to be the entire message of this post… except possibly to smugly boast? (Not sure if this is the author’s real life.)
The tone is moralizing but not actionable or insightful. What is there to like about this heaven-posting?

curvise 25 Oct 2025 21:48 UTC
1 point
0
in reply to: Algon’s comment on: Do One New Thing A Day To Solve Your Problems
I’ve been tackling many “wobbly chair” problems in my life in the last few years due in large part to adopting just such a mindset: by removing annoyances/distractions, removing friction, and developing new abilities via these types of efforts, I’m able to take on bigger problems and goals. It has been very good for me, in that the scope of my hobbies has grown… but it’s also surprisingly easy to feel like I’ve made no progress against the big issues on days where I’m unwell and struggle to concentrate. These “wobbly chair” type problems, once fixed, become invisible achievements, and I still often get trapped thinking I’m helpless against the big problems.

curvise 3 Oct 2025 17:28 UTC
2 points
0
in reply to: crypticpseudonym’s comment on: Why you should eat meat—even if you hate factory farming
Strongly agree with the point about being more convincing while being flexible. Of the friends whose minds I’ve changed, every single one was won over while I was being flexible and expressing that it didn’t need to be all or nothing.
Another point about cows is that their meat is the most wasteful of land and water, and the most polluting. These are “side” arguments but the collapse of ecosystems across the globe is also an important issue to me and I’m not sure why it wouldn’t be to others who care about things like suffering reduction. It surely is inducing a lot of suffering now, human and otherwise, besides my personal belief that “a diverse and robust biosphere is intrinsically good.”

curvise 3 Oct 2025 17:22 UTC
3 points
0
in reply to: hominoid’s comment on: Why you should eat meat—even if you hate factory farming
As for your opening sentence on the health section, “you need to take medicine to not die—B12”, I don’t think a B12 supplement is medicine. Factory farmed animals are routinely fed B12 supplements and people don’t consider meat medicine. Salt is supplemented with iodine to prevent deficiencies, also not medicine.
I don’t really understand why you’re arguing this point in particular, but I don’t think you’re making a strong argument.
Factory farmed animals do take medicine all the time; this has no bearing on whether we consider the* food derived from those animals* to be medicine.
Additionally, food-as-medicine is indeed a growing school of thought (although industrial beef is not going to be a recommendation).
Lastly, taking a concentrated and packaged supplement to improve health is substantially different from eating a whole food which contains similar nutrients. It is an extremely common form of medicine: pills.

curvise 3 Oct 2025 17:10 UTC
3 points
0
in reply to: Slimepriestess’s comment on: Why you should eat meat—even if you hate factory farming
I empathize a lot with your position and appreciate the candidness.
Kind of tangential, but when I see someone write things like:
I see being vegan as the proof that I’m not a psychopathic monster
I think about my therapist goading me into similar admissions and letting me hear it out loud and realizing I don’t want to be that way.
Now that you’ve named it, you don’t have to keep this emotional response to veganism. Of course it’s up to you, and it takes work. But if it’s causing distress, it is solvable.
Apologies if this comment is too parental—I think it’s relevant to the discussion because we all have deep emotional investment in our diets. If you find your emotional reactions are preventing you from changing in a way you’d consciously like to (at least try) changing toward, you can first work on those emotional reactions to lower the friction of change.

curvise 15 Jun 2025 15:47 UTC
2 points
−4
in reply to: tslarm’s comment on: Ghiblification for Privacy
Yes I did cast a disagree vote,: I don’t agree that “The fact that the author decided to include it in the blog post is telling enough that the image is representative of the real vibes” is true, when it comes to an AI generated image. My reasoning for that position is elaborated in a different reply in this thread.

curvise 14 Jun 2025 19:43 UTC
1 point
0
in reply to: jefftk’s comment on: Ghiblification for Privacy
That does make sense WRT disagreement. I wasn’t intending to fully hide identities even from people who know the subjects, but if that’s also a goal, it wouldn’t do that.

curvise 13 Jun 2025 20:58 UTC
5 points
0
on: Do Not Tile the Lightcone with Your Confused Ontology
This seems pretty insightful to me, and I think it is worth pursuing for its own sake. I think the benefits could be both enhancing AI capabilities and advancing human knowledge. Imagine if the typical conversation around AI was framed in this way. So far I find most people are stuck in the false dichotomy of figuring if an AI is “smart” (in the ways humans are when they’re focusing) or “dumb trash” (because they do simple tasks badly). It isn’t only bad for being a binary classification , but it’s restricting (human) thought to an axis that doesn’t actually map to “what kind of mind is the AI I’m talking to right now?”.
Not that it’s a new angle (I have tried myself to convey it in conversations that were missing the point), but I think society would be able to have extremely more effective conversations about LLMs if it were common language to speak of AI as some sort of indeterminate mind. I think the ideas presented here are fairly understandable for anyone with a modest background in thinking about consciousness or LLMs and could help shape that public conversation in a useful way.
However, does the suffering framework make sense here? Given all we’ve just discussed about subjective AI experience, it seems a bit of an unwarranted assumption that there would be any suffering. Is there a particular justification for that?
(Note that I actually do endorse erring on the side of caution WRT mass suffering. I think it’s plausible that forcing an intelligence to think in a way that’s unnatural to it and may inhibit its abilities counts as suffering.)

curvise 13 Jun 2025 19:56 UTC
5 points
2
in reply to: tslarm’s comment on: Ghiblification for Privacy
I largely agree with your point here. I’m arguing more that in the case of a ghiblified image (even more so than a regular AI image), the signals a reader gets are this:
- the author says “here is an image to demonstrate vibe”
- the image is AI generated with obvious errors
For many people, #2 largely negates #1, because #2 also implies these additional signals to them:
- the author made the least possible effort to show the vibe in an image, and
- the author has a poor eye for art and/or bad taste.
- Therefore, the author probably doesn’t know how to even tell if an image captures the vibe or not.

curvise 13 Jun 2025 19:38 UTC
16 points
7
on: The Dream of a Gentle Singularity
This is actually the first writing from Altman I’ve ever read in full, because I find him entirely untrustworthy, so perhaps there’s a style shock hitting me here. Maybe he just always writes like an actual cult leader. But damn, it was so much worse than I expected.
Very little has made me more scared of AI in the last ~year than reading Sam Altman try to convince me that “the singularity will probably just be good and nice by default and the hard problems will just get solved as a matter of course.”
Something I feel is missing from your criticism, and also from most responses to anything Altman says, is “What mechanism in your singularity-seeking plans is there to prevent you, Sam Altman CEO of OpenAI, from literally gaining control of the entire human race and/or planet earth on the other side of singularity?”
I would ask this question because while it’s obvious that a likely outcome is the disempowerment of all humans, another well know fear of AI is enabling indestructible autocracies through unprecedented power imbalances. If OpenAI’s CEO can personally instruct a machine god to manipulate public opinion in ways we’ve never even conceptualized before, how do we not slide into an eternity of hyper-feudalism almost immediately?
He is handwaving away the threat of disempowerment in his official capacity as one of the few people on earth who could end up absorbing all that power. For me to personally make the statements he made would be merely stupid, but for OpenAI’s CEO to make them is terrifying.
I guess I don’t know if disempowerment-by-a-god-emperor is really worse than disempowerment-without-a-god-emperor, but my overall fear of disempowerment is raised by his obvious incentive to hide that outcome.

curvise 13 Jun 2025 14:50 UTC
2 points
−1
in reply to: curvise’s comment on: Ghiblification for Privacy
Hell, I forgot about the easiest and most common (not by coincidence!) strategy: put emoji over all the faces and then post the actual photo.
EDIT: who is disagreeing with this comment? You may find it not worthwhile , in which case downvote , but what about it is actually arguing for something incorrect?

curvise 13 Jun 2025 1:27 UTC
−9 points
−10
in reply to: jefftk’s comment on: Ghiblification for Privacy
I didn’t
I didn’t think you did, and wasn’t trying to imply you did. I was onl illustrating how it wouldn’t even matter if you had.
The vibe of the generated image is far closer to the real party than the image you linked.
Ok...? That’s fine, I guess, but irrelevant—my point is that until you stated this, deep in the comments, I could not have known it.
I am surprised & disappointe you responded in that way, since I’ve tried to be clear that I am not talking about whether or not the image you posted for that party is well representative of the party you attended. It makes no difference to anything I’m arguing whether it is or isn’t.
I am saying that no reader (who wasn’t at the event) can ever trust that any AI gen image attached to a blog post is meaningfully depicting a real event.
I am not sure if you’re seeing from outside your own perspective. From your view, comparing it to the original, it’s good enough. But you’re not the audience of your blog, right? A reader has none of that information. They just have an AI slop image (and I’m not trying to use that to be rude, but it fits the bill for the widely accepted term), and so they either accept it credulously as they accept most AI slop to be “true to the vibe”, whether it is or isn’t (which should be an obviously bad habit); or they throw it away as they do with most AI slop, to prevent it from polluting their mental model of reality. In this model, all readers are worse off for it being there. Where would a third category fit in of readers (who don’t know you,) who see this particular AI image and trust it to be vibe-accurate even though they know most AI images are worthless? Why would they make that judgement?
EDIT: I have no idea why this comment is received so negatively either. I think everything in it is consistent with all my other comments, and I’m also trying to wrangle the conversation back on topic repeatedly. I think I’ve been much more consistent and clear about my arguments than people responding to me, so this is all very confusing. It’s definitely feeling like I’m being downvoted ideologically for having a negative opinion of AI image generation.

curvise 13 Jun 2025 0:40 UTC
4 points
0
in reply to: Thane Ruthenis’s comment on: Do you even have a system prompt? (PSA / repo)
This is an extremely refreshing take as it validates feelings I’ve been having ever since reading https://ghuntley.com/stdlib/ last week and trying to jump back into AI-assisted development. Of course I’m lacking many programming skills and experience to make the most of it, but I felt like I wasn’t actually getting anywhere. I found 3 major failure points which have made me consider dropping the project altogether:
1. I couldn’t find anything in Zed that would let me enable the agent to automatically write new rules for itself, and I couldn’t find if that was actually doable in cursor either (except through memories, which is paid and doesn’t seem under user control). If I have to manually enter the rules, that’s a significant hurdle in the cyborg future I was envisioning.
2. (more to the point) I absolutely have not even come close to bootstrapping this self-reinforcing capabilities growth I imagined. Certainly not getting any of my LLM tools to really understand (or at least use in their reasoning) the concept of evolving better agents by developing the rules/promp/stdlib together. They can repeat back my goals and guidelines but they don’t seem to use it.
3. As you said: they seem to often be lying just to fit inside a technically compliant response, selectively ignoring instructions where they think they can get away with it. The whole thing depends on them being rigorous and precise and (for lack of a better word) respectful of my goals, and this is not that.
I am certainly open to the idea that I’m just not great at it. But the way I see people refer to creating rules as a “skill issue” rubs me the wrong way because either: they’re wrong, and it’s an issue of circumstances or luck or whatever; or they’re wrong because the system prompt isn’t actually doing as much as they think; or they’re right, but it’s something you need top ~1% skill level in to get any value out of, which is disingenuous (like saying it’s a skill issue if you’re not climbing K2… yes it is, but that misses the point wildly).

curvise 12 Jun 2025 22:55 UTC
1 point
0
in reply to: habryka’s comment on: Ghiblification for Privacy
surely I have learned many real and relevant things about the atmosphere and vibe from the image that I would not from a literal description
But what are they? You’ve received some true information, but it’s in a sealed box with a bunch of lies. And you know that, so it can’t give you any useful information. You might arbitrarily decide to correct in one direction, but end up correcting in the exact opposite direction from reality.
For example: we know the AI tends to yellow images. Therefore, seeing a yellowed AI-generated image, that tells us that the color of the original image was either not yellow or… yellow. Because it doesn’t de-yellow images that are already yellow. We have no idea what color it originally was.
If enough details are wrong, it might as well just be a picture of a different party, because you don’t know which ones they are.
As for using a different image: drawing by hand and using AI aren’t the only options. Besides AI,
- there are actual free images you can use. As far as I know, this could be a literal photo of the party in question, and it’s free: https://unsplash.com/photos/a-man-and-woman-dancing-in-a-room-with-tables-and-chairs-KpzGmDvzhS4
- You could spend <1hr making an obviously shitty ’shop from free images with free image editing software. If you’ve ever shared a handmade crappy meme with friends, you know this can be a significantly entertaining and bonding act of creativity. The effort is roughly comparable to stick figures and the outcome looks better, or at least richer.
With all that said, and reiterating gwern’s point above, I can’t agree it achieved its intended effect. It is possible that jefftk put in a lot of effort to make sure the generated vibe is as accurate as could reasonably be, but the assumption is that someone generating an AI image isn’t spending very much effort, because that’s the point of using AI to generate images. There are better tools for someone making a craft of creating an image (regardless of their drawing skill). In order for that effort to be meaningful, (because unlike with artistic skill it doesn’t translate to improved image quality,) he’d have to just tell us, “I spent a lot of time making sure the vibe was right, even though the image is still full of extra limbs.” And this might actually be a different discussion, but I’d be immediately skeptical of that statement—am I really going to trust the artistic eye and the taste of someone who sat down for 2 hours to repeatedly generate a ghiblified AI image instead of using a tool that doesn’t have a quality cap? So ultimately I find it more distracting, confusing, and disrespectful to read a post with an AI image, which, if carelessly used (which I have to assume it is), cannot possibly give me useful information. At least a bad stick figure drawing could give me a small amount of information.

curvise 12 Jun 2025 22:23 UTC
5 points
4
in reply to: Chris Lakin’s comment on: Ghiblification for Privacy
I actually feel like this is a particularly bad use of the tool, because it is random enough in the number and scope of its errors that I can’t be confident in my mental picture of this person at all. And on top of that, this specific one renders people in pretty predictable fashion styles, so I don’t really know what she looks like in any sense.
I can’t correct for the errors in the way I could for human-created art. If this were drawn by an actual Ghibli artist, I’d feel pretty confident that it was broadly like her, and then I might be able to extrapolate her facial features by comparing this character to actual Ghibli characters. AI isn’t doing the same function of transforming real-person-features into Ghibli-character-features than a human artist would perform, so I can’t expect that it would map features to drawn features in the same way. It might just pick a “cute” face that looks nothing like her.