RSS   Atom

Slightly reducing the sloppiness of AI generated frontend

2026-06-12

I've been trying to make progress on a herculean task; I want decent looking programs for personal use which I can generate quickly using AI agents yet I'm a person without taste controlling an AI without taste. I found "one weird trick" that does enough for me, the results are not extra nice but it doesn't make me gag and I'll take what I can.

Style of Slop

I took a simple single web page app (more context below) and kept on asking the agent1 to generate it in different styles. I found that there can be many different styles, all of which feel like slop. Slop is not a distinct style, it can be overlaid on top of many others. Even when I got it to make a page to look like X, it looked like X with slop.

Qt

Only one generation stuck out to me. Simply asking it to make it look like a Qt app - to my tasteless eyes - removed almost all feeling of slop. You can check some of the results out here. This is obviously subjective, hard to convince anyone of and I'd like feedback. I'd also like to see more experimentation in this area, are there any other styles or design guidelines which an AI can generate without adding slop on top?

App context

I read this article from Axios -- a forecast of how the electoral college will change in 2030 and I wanted to visualize it. I dumped a picture in to chatgpt and asked it for a 270-to-win style program with the forecast changes so I can see which routes open or close for R's vs D's.

Other apps

It does translate well outside of this example. I've now gotten codex to translate all the other personal software I have to "Qt style" and I think they look much better.

  1. gpt-5.5-thinking in codex cli ↩

The CIA believes everything

2026-06-07

The Wikipedia page on James Vicary (The "inventor" of Subliminal advertising) mentions that the CIA investigated subliminal advertising and found it to be somewhat useful. After reading the source I found that the report is more skeptical than presented but it did tickle something in my brain; the CIA was (is?) interested in so much bunk! I've compiled an incomplete list of some pseudo-science the CIA was involved in. Lest you think they were just doing their duty and investigating all possible ways to get an edge, I will quote some of their findings where they seemingly 'fell for it'

List of involvements

Uri Geller

After experimenting with him for eight days the CIA concluded

As a result of Geller's success in this experimental period, we consider that he has demonstrated his paranormal perceptual ability in a convincing and unambiguous manner.

source

Dowsing

Maybe prof. Calculus was right after all! From the Discussions and conclusions section of a study on dowsing.

This is the third year in which a computer-assisted search experiment has provided evidence that psychic functioning may be of some use in meeting the military requirement of searching for hidden or lost targets. Although such functioning is not completely predictable, it appears to be robust enough, when selected subjects are used, to significantly reduce the average search time from what it would be if randomly located starting points were used.

"Interesting" theories of consciousness (Do your own research!)

In a 30 page document investigating the "Gateway experience" and hemi-sync, one can find all kinds of great paragraphs. Here's a choice quote.

According to the theories of Karl Pribram, a neuroscientist at Stanford University and David Bohm, a physicist at the University of London, the human mind is also a hologram which attunes itself to the universal hologram by the medium of energy exchange thereby deducing meaning and achieving the state which we call consciousness

Robert Monroe was the inventor of Hemi-Sync. If you want a more complete picture of his worldview, this is from his Wikipedia page

Loosh was a term Monroe coined to describe a type of 'energy' created by humans (and other living beings) experiencing intense fear, grief, or despair—Monroe argued non-human beings who control reality harvest this energy and feed on it.

Misc

I will not get drawn into MK Ultra, but there's a lot more thataway. I will add in passing that the CIA requires a polygraph screening for all employees until today despite all the research showing it to be ineffective at telling the truth from lies1

What to make of it

On the one hand you'd like your government to research anything that might help them defend you. It wouldn't be as interesting if they investigated various forms of pseudo-science and concluded that it will not serve to advance the interests of the state. But they didn't do that, they found a lot of this useful.

What I think may be happening is that the Agency wanted to study a specific topic to find if there's any there there. There aren't many scientists capable and willing to do these experiments. The few that are, are people on the fringes, people who have risked their reputation and possibly destroyed it. You have to assume the reason they did is because they do believe in the theories being tested. Maybe a combination of belief plus the potential for more money for future experiments caused them to design faulty experiments.

If you have any better theories, let me know.

  1. It's possible they just want to test the performance of a prospective employee under pressure ↩

Find windows lockscreen images

2026-05-27

If you see a nice picture on the Windows lockscreen and want to save it;

  1. Open file explorer and paste %LocalAppData%\Packages\Microsoft.Windows.ContentDeliveryManager_cw5n1h2txyewy\LocalState\Assets
  2. Either a) Open each file with VLC1 b) Copy the files to somewhere else and append .jpg

Voila

  1. I believe VLC uses magic numbers for detecting format and Win media player seems to just use the extension ↩

Linux on Macbook Air

2026-03-20

I installed Linux Mint on an old Macbook Air for a relative, here are some notes.

Target machine

Issues

The disk is not visible by default. I needed to enter the option intel_iommu=off in grub as described on super user. The installer did not have good error handling for this and quit with TypeError: "NoneType" object is not subscriptable.

The Broadcom wifi chip needed a proprietary driver. I was happy to find out that the driver can be installed from the install USB, so that was pretty easy. However the driver does not play nice with the current kernel - 6.171 - and I had to pin the kernel to 6.142.

The webcam was more of a mess; cheese couldn't find a device. Turns out "Macbook Air Facetime HD camera" needed a driver too. The driver can be found on github but it took quite a bit of finagling to get it to work.

Enhancements

For a long time Mac user, these helped make the new OS feel more familiar.

Albert, a launcher which looks similar to spotlight. I enabled: Applications, files and system plugins. Had to play around a little with the indexing allow and disallow lists.

Moving the taskbar to the top is an easy little win...

It's 2026, so I described everything I did to the user's Claude. There are no issues yet but that should pay dividends.

Happy Surprises

Microphone, bluetooth, closing the lid / suspend and brightness, volume and backlight keys all worked well out of the box. In fact, wifi is working better now than it did on Mac OS where it couldn't connect to hotspot.

Before the install, the laptop barely worked (opening a browser took ~30 seconds) and felt outdated, now it feels snappy and modern. Maybe I'll update in 6 months to see how it's holding up.

  1. I discovered this myself but it's online too ↩

  2. bash sudo apt-mark hold \ linux-generic \ linux-image-generic \ linux-headers-generic \ linux-image-6.14.0-37-generic \ linux-headers-6.14.0-37-generic ↩

Kharg island

2026-03-15

I have a short term prediction in the Iran conflict.

The US sent a group of marines towards Iran which has raised eyebrows. The US also struck over 90 Iranian targets on Kharg island.

I think the US is planning on taking the island. This fits nicely for a few reasons. First it seems that many of the attacks against the straits have come from there and securing the strait is of utmost importance More importantly, the combined force is still hoping for regime change and so it doesn't want to destroy Iran's economic future as evidenced by the American reaction to Israeli strikes on Iranian oil fuel depots. However Trump has threatened to escalate to attacks against Iranian energy if Iran does "anything to interfere with the Free and Safe Passage of Ships through the Strait of Hormuz". Taking and holding the island with marines enables him to do both - escalate, take Iran's oil, while saving the infrastructure for the future.

~30% by the end of the week

Economics of Continual Learning

2026-03-11

Dwarkesh thinks continual learning is a bottleneck to AGI. Regardless of the technical difficulties, I don't understand how it's ever supposed to be viable for a business.

As I understand it, the definition of continual learning is that the model updates its weights in order to preserve new information (as opposed to keeping memories in an external database - whether that's memory.md or RAG). The idea is that every user would have a model that would learn how to work for that user and would get better over time. That implies that every user has to use a model with personalized weights. This seems impossible to me!

First, there are the obvious storage costs. OpenAI has ~1 billion users and 50 million paying users1. OpenAI would have to store a 1-10 TB model for each one of those users (if they all get access). That comes out to more than $10B a month. Less obviously, all inference providers rely on a trick called batching, which means that instead of running one query through all the layers of the model end-to-end, many queries are processed per layer before moving on to the next layer. This saves time on loading parameters into the GPU. This would be impossible if everyone's models have different weights and would make inference prohibitively expensive.

Yes, a lab can offer this model to a few high paying customers but is that enough of an incentive to spend R&D on? The labs don't even release models RL'd for specific tasks (outside of coding) which would have far more users.

  1. https://openai.com/index/scaling-ai-for-everyone/ ↩

Tmux brackted paste issue

2026-03-02

While perfecting the perfect shade of blue for links, which mostly entailed copy/pasting the values from the default google color picker I came across a (sort of) bug in tmux and simple fix.

When pasting the color hex into nvim inside a tmux window, the hex trailed off with ^[[201~. Weirder was I couldn't do anything anymore, every character I entered was instead represented by the utf code. Even Ctrl-b was hijacked, I couldn't exit (n)vim (that brought back bad memories...), or close the pane. By pure luck, I discovered the copy/pasting something again into nvim gave me back full control.

I found this issue which explains the cause. tmux waits escape-time long after an ESC to decide if the following chars are part of a longer key sequence. Due to lag (or possibly Windows Terminal) and a short escape-time tmux no longer interprets [201~] as "End bracketed paste", it just takes the raw code and passes it through to the open program (nvim).

The solution (as mentioned in the issue comments) is to set a higher escape-time. There is a trade off - esc to normal mode being less snappy - so I'm still experimenting with the right time, but 50ms seems good so far.

I still don't understand why it was messing up my terminal that badly, or why pasting again fixed that...

Further reading

What's the effect of escape-time in Tmux? tmux manual, escape-time