CapsAdmin

My biggest clue is the odd repeating details. Especially on grass, tree leaves, pores, etc

Not all image models (or sampler settings) do this, but with gpt it's very obvious.

With this in mind, I misjudged 3 images on the default game.

This feels very condescending and reminds me of how theists "feel sorry" for atheists because they don't believe in god.

I would say opus was in some ways more accurate, but missed the higher level curvature feel of the site that astra picked up on.

It sounds completely trivial and likely I'm wrong here, but could it be that opus saw the reference image squished? That might explain the sharper horizontal curvature

I've been vary of using ai to search considering all the spam out there. I think I'd rather, perhaps naively, whitelist wikipedia, reddit, arxiv, some news sources, etc than include everything.

Is there nothing out there that does this? I'm paying for kagi and I can see that it has an api, is that maybe sufficient if configured properly?

Is this a normal thing now? I remember seeing this talked about as a surprising thing, but now I'm seeing posts about this as if it's normal.

(I switched to using local models as usage limits, api instability and the concept of paying per token stresses me out)

"Generate an SVG of a pelican riding a bicycle" tests:

https://gistpreview.github.io/?815466e3208746488d47679949b68... - 33170 tokens

https://gistpreview.github.io/?815466e3208746488d47679949b68... - 18125 tokens

https://gistpreview.github.io/?815466e3208746488d47679949b68... - 12960 tokens

Scale 35 felt a bit noisy and incoherent, but 30 seemed nice. (they use the same seed, but idk how reliable seed in llamacpp is)

I use a python test script that captures the answer and renders it to a html page along with the llama-cli log, launch parameters, the chat log, and the python script itself for maximum transparency. :)

This morning I tried experimenting with this ThinkingCap lora I found someone made for 3.6 https://huggingface.co/signsur4739379373/Qwen3.6-27B-Thinkin...

ThinkingCap is a 3.6 27b finetune that claims to halve thinking tokens while maintaining the same output quality. I've used the model a lot and I'd say it holds up. Since 3.6 has the same architecture as 3.8, the lora can be applied.

With the prompt "create a fancy circle in html", these are the results for xhigh, medium, low and xhigh + thinkingcap lora

https://gist.github.com/CapsAdmin/b0ea64006f942c5a96a56dba78...

(Note that the gists are bloated because they contain the full chat and launch params in text/plain script tags for transparency)

I'd say xhigh looks a little better than xhigh + lora, but the lora variant has 40% less thinking tokens. Both seemed to take the same approach with adding random details that weren't explicitly specified.

Medium and low (no lora) are close to each other but are much simpler results.

This is just me testing a single turn. I haven't tested this on multi turns and whatnot, but I thought the result was interesting enough to share anyway.

It comes with MTP already, they just don't mention it in the filename as it's now the default.

I've tracked it for a few years with whatever the latest samsung watch is, and it usually says it's good to excellent.

My total sleep time is about 6 hours. deep sleep is about 1 hour and 30 min. rem sleep is about 1 hour and 20 min.

I usually go to bed around 11pm and wake up at 6 am (give or take, I'm married)

It does however warn me about bad sleep if I've been drinking, have a fever, traveling across timezones, or staying up very late for some reason. One of these things may happen every 2 months or so.

I feel mentally good, I'm in shape and in general good health according to yearly checkups. I've been drinking coffee for about 18 years now.

Whenever I stop drinking coffee for a few days or even months, I find that it becomes harder to wake up. I seem to sleep a lot heavier, but in the morning it's difficult to open my eyes as they feel heavy.

It doesn't matter if I have 1 cup in the morning or drink coffee all day. As long as I have it in my system from the previous day, it's very easy to wake up the next day.

Something that bugs me about these new demos is that they benchmark the model and harness at the same time. The harness steer the model to expand the vague prompt, self correct with tools, etc.

I think it's useful when evaluating "model+intended harness", but I'm more interested in seeing raw model improvements than harness improvements.

That was pretty much my takeaway from reading the website. Not my cup of tea, but I guess it might valuable to some?

I've had to use the aws cli at work a few times (not an expert), but I could see aws trying to compete here if this becomes a thing.

I think nothing can be done anymore, but I don't buy that security through obscurity never worked in practice. A better way to look at this is effort rather than obscurity. The effort it takes to do something with AI is going down, not up.

Almost everything was possible given you put in the effort, but few people possess the will to put in the effort AND pursue a malicious goal.

I think an obvious example is all the fake ai content flooding the internet made to trick people in exchange for money (ad revenue, scams, likes, etc). This existed before ai, but I think it's fair to say pumping out content now requires less effort than it did before.

Most physical locks are an example of security through obscurity/effort. You can after all just pick a lock if you go through the effort to learn the skill. But once a universal lock picker is made available to everyone, you will simply see more locks getting picked.

Lua has the same problem with local functions, and Lua was the first language I learned that I still use today. So it feels very natural for me to see local functions with no dependencies at the top, and the ones with the most dependencies at the bottom. (but you can work around this in lua with globals, forward declarations, functions in tables, etc)

So I guess I kind of prefer seeing helper functions at the top and the main logic at the bottom most of the time.

> What makes someone a script kiddie or not a script kiddie?

The way I see it, a script kiddie is someone who buy weapons with the intention to cause harm. They don't know or care how the weapon works, as it's just a tool for their malicious plan.

A non-script kiddie would be someone who build weapons.

If you assume weapons have recreational purposes, the builder/non script-kiddie is less likely to be malicious as they love the craft of building weapons. For example recreating historical weapons, or just being nerdy about machinery of a gun works.

A builder with malicious intent as first priority can exist, they are more likely to just be interested in building.

So basically, nerds are perceived to be more ethical than non nerds.

I'm of the opinion that AI represents the average creativity, which can be interesting (especially base models) but on the other hand, too many cooks can ruin a dish.

It's also somewhat considered lame and cliche to ask an artist where their idea came from, and what their intention was.

However people will claim that the artist had an intention, it was just not obvious to the artist at the time. The art was also shaped by their experiences, etc. So in contrast, AI does not have this intention.

Maybe we care about where something comes from, but we don't have the will to care about everything? (also, never meet your heroes comes to mind regarding the endless debate)

For example, I had never thought or cared where the trash can in my office came from. I forgot where I bought it from. Now that I'm thinking about it, I'm a little curious, but is it worth knowing? If I knew the designer of the trash can was a bad person, should I get get a new one and make sure the designer of that trash can is a good person?

I think logically and ethically yes, but I'm not willing to spend my time on this.

It's an extreme example, but I think the same applies to music, art, movies, etc, just to a lesser degree for people who claim they care more about the outcome than the human author.

I think most people care about the final output, especially the market. This is why many artists complain about mainstream art, because it genuinely isn't interesting from an artist's perspective.

I'm not too familiar with visual arts, but I know most jazz musicians make a living off of teaching music and doing concerts where the attendees are just other musicians. It's very much centered around technique and human performance. (the term musician's musician is also used here)

Meanwhile, the average person think jazz is noise.

So my point is, saying "What makes music meaningful..." sounds more like an elitist jazz musician take. (I'd say jazz musicians tend to be more self deprecating than elitist, and often make light of the genre and how people perceive it!)

In these discussions, the concept of perceiving content gets intertwined with how content was made, because artists and creatives are usually more interested in how something was made.

So unless these two concepts are separated, people will endlessly talk past each other assuming both sides hold the same fundamental beliefs.

Claiming you made something by yourself, when in reality someone else made it (AI) is easy to frown upon.

I think if you can have a positive experience from looking at a sunset, hearing birds sing, etc, it should technically be possible to have a positive experience from the outputs of an AI without human input. This assumes art is clearly defined as needing to be made by a human with some effort, which neither a sunset or AI outputs are.

In practice, people who use AI to make content give it some direction. The amount of guidance varies wildly, and in practice the majority of what we see is minimal involvement.

I think then it's pedantry. I can kind of see how not having an opinion about what to wear tommorow is by itself having an opinion on what to wear tommorow, but that's not what people usually mean when they say they don't have an opinion about something.

For certain viewpoints that are fundamental ly hard to challenge, a reasonable middle ground to me is to just talk about something else and respect other peoples viewpoints. This is usually how religion is (or should be) treated in open public spaces. But it's by no means easy and I can't really blame vegetarians for being passionate about it either. But I see this more from the perspective of how to behave around other people and not so much what I personally believe, but maybe that's largely cultural. (Norwegian)

For example, I have strong opinions about people believing in things without sufficient evidence, but unless I'm in the correct space or is invited to, I'd rather keep it to myself.

> is happy with whoever rules whom and how justly

This is an assumption. How about neutral, or even hopeless? There's also the position of just not caring because you think it doesn't affect you.

It feels a bit like saying anyone who does not want to fight or talk about a war are on the enemies side.

> You can turn off the non-technical comments by ignoring them.

This also feels like an odd thing to say. If you're going out with a bunch of friends to eat out, and half of them always talk about how immoral it is to eat meat amongst themselves while you're eating meat, I think that would affect me negatively at some point.

One interesting workflow I've seen is that the project maintainer simply rewrites and implements the pull request themselves and closes the PR.

LuaJIT has operated this way since 2012, though with a thanks and mention in the commit message. It seems like a good way to filter out people who prioritizes leveling up their github profiles.

Something a little bit similar, when I was hosting a social game server we had mods. And players always beg for mod status. At first I tried naming the admin group something weird like sandals, but eventually people would ask if they could be sandals too.

What worked best in the end was just hiding it completely making regular players see mods as other regular players. (mods would see who is a mod though)

I would also personally never make someone who asks a mod as it's almost always a sign of wanting power for the sake if it. I would instead just passively observe behavior until I trusted the player and make them a mod. I would then tell them that I don't expect them to exercise their power, but would demote if I see abuse of power.

I'd say it's objectively true to say that open source software is easier to audit compared to closed source software, which you can extrapolate to mean that it's less prone to malicious code injection.

It's not perfect, but surely it's easier to audit for malicious code than closed source.

Also, there is no shortage of volunteers looking out for code changes in established open source software. I think it's fair to exclude software that is very new and/or that has no users, which may be closer to equal footing with proprietary software.

Even for established proprietary software, you get volunteers watching out for changes in releases. Though, far less than open source, and more reserved for people who know reverse engineering.

This is my takeaway as well. Having the source code open makes it auditable, if not by you, maybe the community.

The free software license specifically gives the software an extra advantage in that changes to the software must be shared openly, if distributed as as binaries.

unless it has video input, i wonder if something based on animation and timing would work, as screenshots wouldn't clearly capture motion and response time would be too slow as well

I'm not anti-AI but something I've been thinking about is the discipline it requires. As you said, it's a tool that allows you to rename a variable name on one end and do complete vibe coding on the other end. Developers may say that we should stay somewhere left on that spectrum, because that's where human's are more involved.

But developers also say good practices should be followed when talking to each other, and while some may do, reality is often very different.

It requires discipline, which varies a lot between developers, between projects, current mood, and so on.

In the beginning you might be careful doing small changes, but after a while you might get more tempted to accept the output for what it is, because ultimately that's much easier.

So the way I see it; the left side is harder work and potentially bigger but delayed dopamine hits, the right side is quick dopamine hits. How do we (at least those who struggle with discipline) resist just slipping to the right?

I started out carefully myself and slipped more into vibe coding, but I don't feel particularly proud of it for some reason.

Just to steelman the GP; some people in the company made a choice while the rest had no say.

I personally think the owners should get to decide, but it's an interesting duality.

(assuming it's not like everyone has a share or something, in which case they would've all had to agree I guess)

It seems to me that since the advent of image generators, art has been firmly defined by artists to mean that it was made by a human. But there might be a spectrum of human involvement where the less a human is involved the less it's art.

What happens too often during these discussions is that someone who writes "make me a cool image" gets conflated with someone used ai to fixup a small rock in their natural landscape drawing. (two extreme ends)

One problem though, is that we don't really know how much the supposed human author was involved in the piece. Now that it's becoming hard to judge, people against ai art can proudly change their opinion on on a piece once they learn that it was made by ai. I've come to think this is somewhat respectable, like you see a video of some extraordinary event (before ai) and then you learn that it was fake, just for views or something.

But on top of all this, there are different ways to "consume" art. Artists may think more about who the artist is as a person and what they felt when they made the piece, while non-artists may just enjoy the piece for what it is, detached from the artist. These two perspectives clash a lot.