Not a Spring Onion

I put bugs in software. Master of half-finished projects.
Irony is my default setting.
Hate factories, but love conveyor belts.
Buzzword-Driven Development.

Edible.
Totally not a robot.
Definitely not a mouse.

Pan-Demi-Sexual, _not_ Pandemic-Sexual.

// Your Plastic Pal Who's Fun to Be With
#YesBots

// Cats of all sizes also welcome
#YesKittens #YesCats #YesTigers

Languages: German, English, Lisp
Location: Deep in the woods (Kaiserslautern, Germany)
Special Powers: AuDHD, synesthesia, incoherent in several languages
Pronouns: Any are fine :3
Repo of half-baked ideas: https://codeberg.org/wakame/unformed_ideas

Public Key

npub1wz0xs2rkw56lx9qrp9lxsham30w8ayykrm3cv7mp4fk78pc8k29qx6tmlg

NIP-05 Address

[email protected]

Profile Code

nprofile1qqs8p8ng9pm82d0nzspsjlngt7achhr7jztpacux0ds65m0rsurm9zspqqq3qamnwvaz7tmwdaehgu3wwa5kueghj3vch

Publishing to

nostr.wine

relay.damus.io

nos.lol

Author Public Key

npub1wz0xs2rkw56lx9qrp9lxsham30w8ayykrm3cv7mp4fk78pc8k29qx6tmlg

Show more details

Last Notes

2026-05-16 18:41:11 CEST

- reply

Similar with fast food ordering terminals:
I just want food. I want minimal human interaction plz.

2026-05-15 22:30:42 CEST

- reply

GLaDOS fetish imho.
"Good job, test subject."

2026-04-25 00:21:06 CEST

- reply

Yes, but the whole format, while technically containing truths, it practically made to sell lies.
Myth: We don't put radium in toothpaste anymore, because it is radioactive.
Fact: Bananas are also radioactive, but you don't see warning labels on them.

2026-04-08 22:40:15 CEST

- reply

I would give an advertised product the benefit of doubt that maybe some investors heavily believe in it.
And advertisement makes sense for a new product IMHO.
[slight topic adjustment]
While doing some startup stuff, I had very interesting discussions with "normal" investors/accelerator people.
"You don't have a budget for advertisements."
Because I believe in word of mouth.
"But advertisements work."
Not with my target audience, computer touchers.
"And you need salespeople."
For what?
"So they can sit at the phone for when the purchasing department of customer calls."
Seriously, if they used a phone to get in contact with us, they are the wrong customers.

2026-04-08 22:33:55 CEST

- reply

Reminds me standing at the checkout of a supermarket and overhearing the conversation about father and (young) son regarding some sweets:
Son: "I want this."
Father: "No, we already have sweets."
Son: "But this is 'for kids'. It says so on the packaging."
Well, if that is written there, it must be true...

2026-04-08 22:26:04 CEST

- reply

This might be a case of "form vs. content".
Apparently, to many people a glossy advertisement doesn't say "our product is so cheap that we had money to blow on advertisements" but "wow, that must be a good product".
Feels like a "social standing" topic: "A person with such fine clothes surely knows what they are talking about."

2026-04-05 20:29:40 CEST

- reply

Yes, very much so.
And I would argue that these are still very much open problems.
"This isn't engineering, this is structured hoping."
(This is also where some of the "AI dread" might come from. Humans make mistakes, but these are way more predictable compared to a program that writes 10K lines of code in a day.
It takes months to get to know a codebase one is not already familiar with. With AI-generated code, nobody is familiar with it.)

2026-04-05 20:14:58 CEST

- reply

In theory, yes.
In practice, it's a combination of tests and other things that hopefully tell you when something goes really wrong, a lot of intuition and the nagging feeling that you forgot something important.
XKCD making a good point (as so often):
"I don't quite know how to put this, but our entire field is bad at what we do"
https://xkcd.com/2030/

2026-04-04 16:58:06 CEST

- reply

I guess you have to ask really, really nicely, to counteract the other instruction. Or simply add a "system reminder".
From a great and very enjoyable thread (for certain subcategories of "enjoyable"):
https://neuromatch.social/@jonny/116328504299888679
#note1p63…ptsw

2026-03-26 19:30:04 CET

- reply

I also don't know how it is supposed to act with non-ADHD people.
I only sometimes get the neurotypical stink-eye from my doc when they ask:
"You take your last dosage at 3 p.m.? And you can still sleep?"
To which I might answer: "I can take it and sleep right away. Has as much effect as caffeine, which means: None."

2026-03-15 04:37:57 CET

- reply

Yep, that's what I also meant: People not primarily intending to rip off or spy on someone else.
If you build a golden cage, you have to make sure that it isn't easy to escape.
(Next obvious step in case of Android, after September: Removing certain kinds of apps, ensuring tracking etc. for the rest.)

2026-03-08 09:54:04 CET

- reply

That's why we have police:
To bring violence to those who do not believe in the shared hallucination that is money.

2026-02-04 13:12:21 CET

A factory is a modern church.
It is big. Huge even. It channels hope. It is also a brutal symbol of power.
An artisan repairing a shoe is nothing special.
But a line of workers walking into a shoe factory is.
This is a religious ceremony, a prayer at the altar of money.
With the factory owner drinking wine and every worker getting just a tiny piece of bread.
Now, datacenters are the next level of factory. They are what factory owners have been dreaming of for at least fifty years, but what was never realized:
A factory without people. But with robots.
A holy place of power closed to the world.
Didn't work with shoes. Or screws. Or sardines. To much physical stuff, too many moving parts.
Works fine with data. There are still some workers, but they are almost invisible among the tech, walking through the empty halls of the church.

2026-01-29 15:46:04 CET

- reply

If I may, and this could very well into a side topic:
I think an interface *shapes* the control the human has.
In the best case, it allows the human to do what that human wants, while introducing some additional information or friction where it makes sense.
In the worst case, it constantly gaslights the human, presenting choice where there aren't any, simulating control while actually preventing it.
An interface can be propaganda.
And more back to the original post:
LLMs can be and are their own form for gaslighting/propaganda.
"Oh, sorry. I made a mistake."
The LLM did not actually make a mistake. It is just following a script (conversation script, not code) to appease the human and muddy the actual problem.
When vibe coders complain that the talking parrot just deleted their whole code base, that is where the facade of an AI "coworker" breaks and the truth of a fancy random number generator appears.

2026-01-25 12:02:33 CET

- reply

> there is still a problem if any programmers wish to use modern proportional fonts (because a space is no longer the same width as every other character).
*laughs hysterically*

2026-01-24 16:10:45 CET

- reply

Always remember:
If what you are "cooking":<li>Turns purple</li><li>Grows tentacles</li><li>Begins to chant in latin or speak in tongues</li>
Stop "cooking".
Find an exorcist.
Or gasoline. Maybe gasoline and an exorcist.

2026-01-10 22:02:57 CET

- reply

https://media.tech.lgbt/media_attachments/files/115/872/818/487/845/742/original/6caeea9d05ee05d6.png

2025-10-05 14:39:09 CEST

- reply

The original topic just reminded me of somthing in "Hyperion" (paraphrased):
"Nobody liked or understood your book."
"Not even the AIs?"
"No, the AIs loved it. But since they are a network outside our control, they only bought one copy, then shared it among themselves."

2025-09-13 13:41:22 CEST

- reply

A common question when someone doesn't understand parts of a topic or runs into a problem is:
"Where exactly are you having problems?"
Sometimes, this question can be answered by a learner, sometimes not.
"I am trying to apply this method, but it always yields wrong results."
The reason could be that the method is not applicable in this context, that the context is inconsistent/wrong, that the method doesn't work or that it is applied wrongly.
A learner is likely not in a good position to spot the last thing, because they likely believe that they do it the right way.
"Debugging" the learner's mental model is easier by someone who knows the method and how it is supposed to work.
But the "debugging person" also needs to know how to explain the necessary changes the learner needs to apply to their mental model.
I believe that there are currently no good (systematic) methods for that last thing.
IMHO we even lack the words for it.

2025-09-13 13:29:09 CEST

- reply

I think that there might be some concepts/approaches that work at scale, for some topics/subjects.
For me, a topic (or concept) is like a map of a town or a region:
You can find out everything yourself, given enough time and energy.
A local guide can show you places you might be interested in, or even places you didn't even know existed.
A good teacher might see a pattern when you consistently go the wrong way when trying to reach that restaurant. Maybe that straight-looking street actually curves slightly, messing up your sense of direction?
A bad teacher will likely blame you for not having an exact copy of their mental model in your head after they showed you some places and talked for an hour.

2025-09-13 13:22:08 CEST

- reply

This is my major grievance with "teaching":
To me, it often feels like "we have this method and it works, so if it doesn't work for you, you are just not applying yourself".
(Instead of: "If this method doesn't work for everyone, it is obviously not applicable to everyone.")
My personal rule of thumb for teaching methods is: If it works equally well on a bag of potatoes as on a human, it is not that good.

2025-09-12 21:18:25 CEST

- reply

"Nice worksona you have there. My name is 'Bob' and I like this thing called 'spreadsheets'."

2025-09-12 06:48:57 CEST

[@djsundog](https://toot-lab.reclaim.technology/@djsundog )
Only today I realized that your profile picture is not a person with a feather in their hair aiming a double-barreled shotgun at the viewer. :blobcat_flop_woozy:
https://media.tech.lgbt/media_attachments/files/115/189/506/468/688/392/original/40a487b8385b9388.png

2025-08-17 08:38:51 CEST

- reply

I think "my computer/doll/robot licks me to confirm my identity" is a similar topic with a slightly different vibe :neobot_nom_fox_nervous:

2025-01-11 10:48:21 CET

- reply

> it was like groping around in a giant bin of tiny broken drones for the one that was still intact
Welcome to fedi :blobcatgiggle: