Categories
Other

Dear Jimmy, please fend off the bots

I have given small, but regular donations to Wikipedia over many years, because I believe Wikipedia provides real value and contributes to the sharing of knowledge and information between human beings. Today I got another donation reminder from Jimmy, and I might as well publish my knee jerk response:

Hello,

I support Wikipedia. But this year I have not yet made up my mind with regard to donations.

Companies like OpenAI, Google, Meta and Microsoft are collectively causing the destruction of the free internet with generative AI technology. The open web made by humans for humans is dying, and it is replaced with machine garbage slop. Communication between humans is clouded by gen-AI-transformations, bots, made up fakes of all kinds and deceptive social media algorithms.

And these companies steal whatever data they can get their hands on without any ethical considerations, solely for their own profit. These are highly unethical companies that cause direct harm to [the] human civilization with their technology and actions.

Before I consider donating to Wikipedia going [forward], I want to know that such a donation does not contribute even the slightest to the enrichment of the big tech companies, by enabling them to continue harvesting fresh data from Wikipedia.

What are you doing to:

  1. Stop AI companies data theft, which is enabling only their own enrichment.
  2. Stop AI slop from degrading the quality of Wikipedia articles.
  3. Stop bots from producing fake and potentially ad biased content on Wikipedia.

Thanks,
Øyvind Stegard

I don’t expect to receive a reply. And I know Wikipedia is already taking actions against abuse, but I think it is such an urgent matter that they should put statements about it directly in all their requests for more donations. Also, they are allowing AI companies API-access through payment, but this is not good enough. AI big tech should be outright denied access.

The world is changing. I am no longer sure I want to directly support a huge source of data for unethical AI companies. Maybe all hope is lost for sites like this, and maybe it is better that Wikipedia and the rest of the web dies a slow death by AI slop, so that the AI bots can harvest and ingest their own excrement to an increasing extent. Bullshit if you like. It’s the same thing. Words, pixels and sound samples computed by soulless mindless mega clusters of hardware. No understanding, no creativity, no consciousness – merely replicating and mixing up the stolen art, communication, written knowledge, original ideas and inherent patterns produced by human beings throughout time.

My current stance is becoming clear: generative AI technology is contributing a net negative to humanity. We would be better off without it. Technology should benefit humans and not the reverse.

And please do not confuse generative AI with machine learning in general. Generative AI is a specific use of machine learning where the goal is to generate data of some kind. To achieve high enough quality, the efforts required to both train and run the large language models are enormous. And so the tech is mostly impossible to democratize and it is not controlled by the people. This gives rise to further power concentration (loss of freedom). And with eternal growth requirements and a winner takes it all profit race, the data, energy and hardware requirements are rising rapidly. Nobody is holding these damned companies accountable for the effects of their actions. The negatives are completely drowning in hype noise and global FOMO.

Instead of focusing all efforts on reducing the climate crisis, ending wars and helping people who suffer we get fucking chat bots with “funny” names (to somehow excuse that they are shit), a slew of vibe coded «fast food software» of mediocre quality, straining of the free software ecosystem with low quality contributions, fake media all over the place, automated propaganda, open web destruction, intellectual skill deterioration, addiction to proprietary LLM services, ad-ified LLM services (did someone say the word «bias» ?), communication sloppification, hardware price increases, electricity price increases, job loss and ultimately loss of freedom. And in that process we are also making the climate crisis even worse. No fucking LLM is going to solve that.

So I’m sorry Jimmy, but unless you completely deny access to the destructive forces of evil and unethical AI companies, Wikipedia will no longer receive donations from me.

R.I.P. Wikipedia and the world wide web that was.

Categories
Other

Mobile only web design

Before, there were no smart phones, and web design was tailored to the computer screen. Then smart phones and tablets happened. Regular old web didn’t work so well on those, so the design became responsive and adapting to device screens. That often works well when someone actually cares about larger screen variants and implements dedicated layouts and styling.

Then mobile first happened.

Design for the mobile first and foremost, then consider designs for larger and wider screens as an afterthought. Too often it is actually mobile only design. Target the phones, and just forget about everything else. (Hell, just forget about the web entirely and force visitors to install an app, mhm ? Fuck that by the way.)

A mobile only web design may visually scale up without becoming ugly, but totally fail to take advantage of a larger screen with precise pointing device and a physical keyboard available. There is often too much wasted space; too much air, too big fonts, too big click (tap) targets and content hidden behind interactive revealing widgets. Too much imagery and visual design crap that actually drains mobile batteries faster. It becomes annoying to visit from a desktop.

In a mobile only web design, the menu is typically not directly visible, because it would take up too much space. So it requires interaction with a hamburger widget of some sort, which will reveal the menu items in huge font, typically covering whatever page is underneath completely. An anemic site map of sorts, because a real site map is just too much content. This is useless on a bigger screen – direct access to all menu items is much better. Similarly, other space saving measures just cripple the user experience on big screen, because of the narrow constraints. See the irony by the way ? The very WordPress theme I use on this site has some of those traits I do not like. (The newer ones are even more horrible, so I’m stuck with the lesser evil.)

Another common space saving measure is the Show more.. widgets. Instead of just laying out all the content and let users naturally ignore, scroll or flick past uninteresting stuff, the interface requires extra interactions while reading. When I’ve opened the next page of a book, I don’t want to be forced to do things with that page to read it. I just want to read it, skim it or skip it.

By all means, it makes very much sense to create mobile designs, but please don’t forget that the web is at its best when accessible on a variety of devices, including *OMG* actual computers with big screens.

I will end this rant by paying tribute to another lovely rant that explains this much more eloquently:
https://motherfuckingwebsite.com/

Categories
Code

ox-tagfilter-js version 1.0

I’ve just released version 1.0 of ox-tagfilter-js, after finally adding a search box to the UI. If you are an Org-mode user and publish org-exported HTML documents from your org-files, you might find this extension useful. It allows you to quickly filter document content, based on tags and heading text, using a web UI.

I use it to lookup stuff in my «digital brain», a growing huge journal of private notes that I maintain as an org-file. The org-file is automatically published to the web, and so I can read and search it easily from anywhere, including with my mobile phone.