bambangan

Rediscovering RSS

Sat, 29 Aug 2026 23:48:43 +0900

From today, a proper RSS feed experience is available from bambangan.net. Not a stripped-down version for search engine bots, but one compatible with reader applications.

I’ve always thought of RSS as some kind of ‘dead’ tech, as I’d never seen anyone use it after the death of Google Reader. But to me it seems to be having a comeback lately; people sick of all sorts of recommendation algorithms are apparently turning to good ol’ RSS to replace their twitter feeds.

So anyway I’m going to implement that into my site. I’m not sure if anyone would benefit from that, but just in case.

As this site is bilingual (Kor-Eng) I’ll keep two separate feeds.

bambangan.net/en/index.xml for English posts
bambangan.net/index.xml for Korean posts

I write in Korean first and then translate to English, so they’ll be the first to get updates. Also, not all posts are translated; just the ones I like.

I’d better try reading with RSS myself as well—just to see how others are doing!

View signed Markdown source

RAG is Busted

Wed, 26 Aug 2026 00:00:00 +0000

About 2 years ago in Korea, ‘AI side jobs’ were a thing. One would create multiple Internet blogs (tistory.com was notorious for this) and then generate a bunch of articles with ChatGPT, monetizing any traffic it attracted. It was so popular that there were even paid video tutorials on ‘automating blogs to get rich quick’. I was quite amazed at the time how much hype it was getting.

And then there’s this rule of economics that any discrepancy between the nominal and real value of an asset leads to people seeking to profit from it. The gap here was caused by ChatGPT dramatically lowering the investment costs of writing a web article (time, effort). With hundreds of machine-written articles and automated tagging systems, these websites tended to do fairly well on SEO, leading to some generating hoards of income using such methods.

But Economics 101 steps in again; the same speculative forces also tend to close the gap between the values. It was not long before those blog ‘content farms’ stopped yielding noticeable profit. Sites that at first were rockin’ with a few banner ads are now bloated with loads of popups and anchor ads. The more these nuisances, the less readers tend to visit. And they’re not without tools! Who would want to visit and read those ad-stuffed internet blogs when they can just send an agent to do it instead? Leading to less traffic, more bots, less revenue. The 2024 ‘blog automation’ craze ended like that.

It wouldn’t have been much problematic if this story was just some lesson of there is no such thing as a free lunch, but it’s far worse than that. To me it seems that these very actors were the ones that changed the future of the web. They’re the ones who killed the Internet.

Lemons take over

Do you remember the 2010s? There used to be a time when people would read up articles found by a Google search and then put together information inside their heads. It sounds really painstaking written like that, but it wasn’t quite. One of the biggest reason is that the quality of information at hand was noticeable at a glance: “Does the site look trustworthy? pass. The author’s profile? seems ok. Are the images scraped from somewhere else?” and such. But today those same heuristics just aren’t relevant anymore.

Now that AIs can build a nice-looking website in ~10 minutes, where on earth would be a crappy website? Image generating models can create author profiles, cover images in a blink of an eye. There are even loads of fake news sites found across the network.

Moreover, it tends to take a lot of effort to ‘detect’ AI-written text. We all have the experience of reading an internet blog when in the middle of reading something starts to feel off: that feeling of awkward phrasing (even when grammar is errorless!) and the feel of AI prose. At that point you realize you’ve wasted ~10 mins of time reading that slop article. When readers have no idea if the content I’m reading is written by a thoughtful human or generated by a single-sentence prompt, they simply decide not to make the bet.

A classic example of a Market for Lemons. The writer of course knows how much effort he put into that article, but the reader has no idea–until he reads it. If in the past the various signaling channels for conveying content quality were operational, LLMs have effectively neutralized much of them. And in a situation of severe information asymmetry, Peaches cannot hold a hand to Lemons. The effort of those who write good content gets undervalued, and they now have no incentive to write online.

A positive feedback loop

garbage in, garbage out.

This phenomenon affects AI power users as well – negatively. Haven’t you had the experience, where you click on that ‘reference’ your LLM provided just to see another AI-written article on the internet?

To mitigate problems with being trained with outdated data, LLMs use a technique called RAG to fetch the recent information. It effectively works like giving them Google search bars; referencing internet stuff to give answers with the newest data.

The main idea about RAGs was to let LLMs use the Internet (or whatever source) as a good source of truth while writing. But who writes those text these days? The concept of feeding GenAI models with its own outputs as training data is known to be a bad idea. While RAGs are not training mechanisms, a similar problem may potentially occur. Especially hallucination! While hallucinations of LLMs itself has decreased substantially over the years, a circulation environment we just built poses a graver problem. AIs, when fed inaccurate information generated by other models, primarily with methods like RAG, would facilitate an environment where false information is created and spread exponentially. Kurzgesagt had a recent video about misinformation spreading through AI slop content being the biggest threat to human knowledge.

So what?

As usual, any novel idea I think of isn’t really novel: a bunch of researchers around the world have already done quantitative works on the topic and written papers about it. And sure they talked about solutions! Here’s an excerpt from the ACM2026 paper Retrieval Collapses When AI Pollutes the Web, written by three researchers from Naver Labs.

(…) as Agentic AI begins to autonomously publish content, defense mechanisms must evolve from static text analysis to behavioral fingerprinting, identifying and isolating agents that systematically produce high-entropy, low-factuality streams.

So we need to change our system, of course. One big issue leading to the AI slop problems of today is that search engines fail to filter them out - a clear limitation of traditional ranking systems. So by reinforcing our search systems, we could be able to put the task of determining whether a text is AI slop or not upon the search engine, not the user.

Anthropic recently announced that they’ll be putting watermarks to Claude output starting August 2026, and it’s not irrelevant to this. Instead of relying on brittle methods like those AI detection tools, we could intervene from the generation process to make AI-written text distinguishable. While this clearly isn’t a perfect solution on its own, it reflects a growing concern that AI writing should be made identifiable.

After the blind optimism with the advent of LLMs went away, we’re now faced with the threats AI systems pose to the human knowledge base. But that wouldn’t mean removing AI entirely – it bears potential and benefits simply too large to ignore. Still, the task of tackling the problems of AI slop and restoring the Internet into a ‘treasury of human knowledge’ would be left for humans to complete.

View signed Markdown source

An AI-dependent World – and what I’d do

Mon, 03 Aug 2026 00:00:00 +0000

We shape our tools and thereafter our tools shape us.
– Marshall McLuhan

More and more people are getting dependent on AI tools for doing smaller things. From writing work emails to choosing what to have for lunch, tasks are becoming harder to complete without the assist of LLMs.

When people think of a dystopian future controlled by Artificial Intelligence, they expect something like Skynet. But if such a future truly comes, I suspect it’d probably be much closer to today than The Terminator.

The outsourcing of thought. As people let machines do the thinking for them, we’ve come to become passive identities not even capable of thinking without any outside help. Where is this trend headed?

Paradoxically, as increasingly powerful automation tools drive the surge in human productivity, humans have become less capable. Consider the number of employees capable of crafting a nice email without LLMs, or the number of developers who could write a JSON parser without code autocomplete (just by reading the API Docs! really?).

So here comes the question of then what. In an increasingly AI-dependent world, what should I do to extend my abilities? Which strengths do I need?

Since then I’ve come up with two different approaches:

A) Learn to work without tools.

This doesn’t necessarily mean ditching every single task-automation or AI productivity tool. It simply means using them only for what I can already do by myself. So everything could stay under my control. This would be both a huge advantage and a disadvantage at the same time – when others crank out their essays at the click of a mouse, I’d be fiddling with subtle expressions or buried in books somewhere in the library. It’d be a hard path. And this option would face the concern that it attempts to run against the advancements of technology.

B) Learn to work with tools.

We have the new technology – why not embrace it? This option inherently entails heightened dependence on these so-called tools; instead of only automating what I could have done without AI, I could start doing things I wouldn’t have been able to! Hugely enhanced productivity is anticipated, but more risk as well.
Not so long ago, everyone was freaking out on Reddit when ChatGPT suddenly went out of service. If something similar happens again? or if inference providers suddenly decide to charge more for their services(or more realistically, phase out their free tiers or usage without logins) - Do I have to pay more? External dependencies bring in all sorts of risks. Understanding those risks and managing them would be a critical part of my job.

The conclusion seems rather obvious; choosing either side (and ditching the other) is too dangerous an option. So I chose to take a hybrid approach: embrace technology, but acknowledge its limitations. To keep my abilities from becoming over-reliant on AI tools, I should train my muscles of active thinking.

I figured I’d start to:

Write. Without the assist of AI, get some time to ponder on some subject for some time, and practice writing on my own. Like how I’m doing right now as I write this article! (This blog maintains a No-AI Policy.)
Read. AI-generated content is bite-sized snippets, but getting used to those would degrade my abilities of finding information by myself. Search for human-written text. Even if it’s a bit long, or a classic. I need this ability to retrieve information from raw material.
Think. There are few things more stupid than staring at my phone while walking. Whenever I stroll around, ride a bus, or have noting in particular to do, embrace daydreaming. Better thoughts need some processing time to emerge.

I feel writing posts like these often quite guide me to see my weaknesses. But it’s much important to contemplate on what needs to be done for self-improvement as well. Especially so as to become one with his unique strengths.

View signed Markdown source

Hello World

Fri, 12 Jun 2026 00:00:00 +0000

Sup.

You're reading my very new Internet blog, which I've just constructed overnight. Yesterday it had occurred to me that I really needed some place like this, and here's what I did with it. I guess the rapid progress was possible partly due to the extensive planning I've done over the months inside my head.

Shall I introduce myself.

But the sole purpose of this website is not to. To share my thoughts without telling my identity. So dear reader, I regret to tell that you'll probably never get to know my name. Still, you would be able to know a great deal of who I am. That's good enough.

On translations

One side note. While I've implemented functionality to add translations like these to my Korean articles, I'll probably won't translate every article. All articles are translated by hand, painstakingly by me (Sorry for the crappy English), since I want to avoid LLMs silently contaminating my words. If you're okay with machine translations, you may head on to the korean version of this site to explore.

View signed Markdown source