LLMs are not only wreaking havoc in nature through their data centers, being heat waves on a planet that is already heating, causing air pollution disrupting the life cycle of living being sensitive to it as well as noise pollution to disturb even those that are less sensitive to light, wasting water, emitting toxic material, not only causing epistemic collapse by encouraging (and pressurizing) people to outsource their thinking leading to atrophy, not only threatening us, normal people existentially, at least by making us economically irrelevant and taking our political relevance as well, leaving us disempowered, alienated and pushed aside (worse cases are enslaving or killing us, which we must also take seriously as possible scenarios), but they are also insecure.
neo: LLM use is impolite
2026-09-06 01:46:25
We are not powerless at all. We’ve got tools.
neo: We need to restore trust in knowledge
2026-09-04 00:34:19
Epistemology is the philosophical study of human knowledge. That is, when we want to think and study knowledge and knowability, then we turn towards epistemology. Human knowledge was an important matter even before epistemology became a field of study and affects basically everything knowledge-related, including, but not limited to education, science, gathering information in order to form a world view or a political opinion. If we have a robust and trustworthy epistemological framework, then we can gather reliable information for further thinking, research, or simply for informed decision-making. We had such a robust framework up until our recent past when LLMs destroyed it, making a new epistemological framework necessary.
neo: A tribute to Hannah Arendt
2026-09-01 03:28:20
Hannah Arendt was a Jewish-German-American historian and philosopher. In 1933 she was imprisoned by Gestapo for illegally researching antisemitism. Upon release she fled to Paris and she later researched on what happened to her beloved continent, Europe.
neo: You are important
2026-08-28 16:08:12
Yes, YOU
neo: Kleptocracy
2026-08-15 18:05:36
Large language models cannot function without training data. They are stochastic prediction systems based on the information they were trained with. Of course, reliability is a separate problem, which is outside the scope of this article, yet, at this space it is suffice to say that whenever they are trained on unreliable (sometimes deliberately poisoned) data, their premise is already mistaken and therefore their conclusion is unreliable. Furthermore, when it combines multiple sources into one conclusion, then, given its lack of understanding of said sources, the result may be complete nonsense. But, again, this is not our topic here. We need to discuss the kleptocracy involved.
neo: Are LLMs worth it?
2026-08-12 19:47:09
We are being told it boosts productivity, it is a force multiplier, it increases profit, etc. etc. etc. So some people are convinced it is worth it for them. We could also debate on that, because speed alone is not a viable metric to determine whether something is worth it. I would be able to make a painting way more quickly than Da Vinci produced Mona Lisa. Would my sloppy way of working be more productive than mister Da Vinci? The answer depends on a value choice and a moral choice: do we value quality or do we value quantity? Good work takes time and even though I could make 100 hilariously bad paintings while the master is working on Mona Lisa, my work would be total garbage in my view, because in case of low quality work the nonexistence of such work is better than its producing in my opinion. It’s better not to eat a soup if it causes diarrhea or worse. And, in my view those whom reduce costs at the expense of quality do a disservice to the world, as they waste resources, lower the expectations and usually market their work as something better than it actually is. The other position that I disagree with is all about minimizing costs without a moral limit of producing at least some acceptable quality.
neo: LLMs in the world
2026-06-12 18:00:00
With the advent of generative AI or LLMs we need to assess what it has taken from us and whether its claimed benefits are worth it. First of all, it scraped the internet and downloaded all publicly available thoughts, ideas and every creative exchange and now, as a tool that absorbed all human expression it generates responses from what it has plagiarized from mankind and sells it back to us as wisdom. Every single image, video or text the AI generates for us is the combination of human expression, ideas and works scrambled together into one result and now the very people whom were the unattributed sources of this giant predictive tool find themselves struggling. Why would one buy a painting or buy other artistic work, such as music, movies or – God forbid – a book when all the wisdom we have ever shared so far is there in the training data and this product can combine the works of our greatest minds into a product of its own? It plagiarizes from unattributed sources and we call it intelligence.