danieltanfh95

There is nothing physics/metaphysics about this. If you don’t understand the terms, don’t pretend you do and write slop as a comment, it is really not that different from using LLM to generate slop.

This isn’t about whether the writer uses LLM or not at all, nor is it about respect. The core novelty it tries to introduce is not hard to understand (even if it is not really that novel). If you don’t want to spend time thinking about what interesting idea it is exploring, that is fine, but pretending or insinuating that it is a LLM problem is just lazy.

As I understand it, content addressing function content is problematic because it does not actually "normalise" the content of functions into something interchangeable. A function of input A and output B with performance signature X can still be very different in terms of actual code, but the actual comparison between both is hard to specify.

I was exploring this as a means to solving the open source, or rather the github conundrum, the problem of sharing code socially is that we need a canonical source, and this is sociologically driven than performance driven, and as it turns out, have devastating consequences for FOSS funding. I wanted to explore sharing code "interchangeably" in some sense to avoid this problem, but ultimately this seems unsolveable, even with exploration by Unison etc.

The article is a farce. Is this really the sort of slop we want to use as a proof that humans write better articles than AI?

> "If a company builds a machine that, when fed descriptions of assorted ethical dilemmas, emits sentences either of the form “Compromise your values” or “Don’t compromise your values,” it is not building a tool that assists people in their decision making; it is encouraging people to stop making decisions. "

A human is not diminished by access to tools or other humans.

As much as we want to pretend that decision-making is what makes us human, the economy and governments are built on delegation. Choice paralysis is a thing.

There is so many logical fallacies in the article I don't even know where to begin.

Not all bots are bad, and the economic incentive of playing nice in a long lived session bot is much more stronger otherwise, which is kind of the point. It is the same with humans.

It is really not that difficult.

1. For juniors, any sort of proof you have passion is enough. 2. Treat mid-levels as seniors. 3. Seniors have to show proof of passion (with longevity and intensity, AI and consultants exist for expertise and menial work) and competence (Take home problems with a long time scale with whatever tools they need. Use AI. I don't care, but be expected to be scrutinized for your design decisions, depth of exploration, and architectural write-ups.)

This is why we need interviews to filter off people who are bad enough to think "bar exam" is the right abstraction for software engineer.

The real solution is stop proposing solutions like a bar exam. Software engineering evolves too quickly for anything but a work sample to work.

It heavily discourages bot farming, which is what makes bots economical.

> This is leading to a class distinction in writing: anything the audience classifies as genai becomes low-status. And the reverse: writing that the audience classifies as human becomes higher-status.

I see this to be the exact thesis that my article was trying to convey, though someone else put it more succinctly:

> I guess the irony is that dismissing points of view about how things could be improved here because they aren't sufficiently dogmatic is pretty much exactly what I understood the thesis of your article to be.

> I guess the irony is that dismissing points of view about how things could be improved here because they aren't sufficiently dogmatic is pretty much exactly what I understood the thesis of your article to be.

Exactly, yep!

This is pretty much my writing with some edits around content I wanted to simplify. The article linked is not the same content at all. Not only does it only focus on the American perspective, it proposes the opposite of what I was trying to convey. The idea that this article is anywhere the same as "read an article containing essentially the same information without the pesky LLM voice" is just flat out misleading.

People claimed I was writing a propaganda piece apparently, when I note that this was a failure of losing empiricism and pragmatism that America had, as a philosophy since 1878 that propagated into the New Deal, because of the Soviet threat.

The counterfactual was not to imprison him in the first place. That was what the orignal article : https://danieltan.weblog.lol/2026/05/americas-greatest-strat... is about.

Qian was widely considered as a "strategic" scientist in China. The knowledge he had counteracted the Soviet political apparatus, and was enough to propel China into superpower status in less than a hundred years.

Exactly what I was trying to convey.

If people are wondering why the linked article doesn't contain the movie reference, the original article was https://danieltan.weblog.lol/2026/05/americas-greatest-strat... which specifically mentions that this is more than a "blunder" but a loss of empiricism and pragmatism during the McCarthy era, along with various caveats and historical detail around Soviet and Chinese attitudes around the methodology Qian brought.

Wikipedia alone does not contain these information.

The article is a sensational piece without a large historical caveat on the Soviet response to Qian moving to China, along with the very important information that his wife was KMT, and they were treated as traitors to the Taiwanese government at the time. However Qian was insistent on removing any trace to America and he refused to visit America thereafter.

I considered against publishing this piece because it could be taken as propaganda.

But America did derive the same idea, back in 1878 and was part of philosophy behind the New Deal. So what do Americans have to be sour about and call it propaganda for, when the American academia and government abandoned empiricism for whatever they believed in 2026?

I don't really think there's a tangential detail that is related to the message. Which one are you referring to?

Also, the upbeat and persuasive style ... is my style kek, is it me being too pushy or?

Unfortunately there is too much detail here for me to write more candidly.

Make your customer, sales and product team prototype their idea with ai directly against the codebase or website before handing it to the dev team to filter off inconsistencies. Reduces communication costs.

yep, as I was exploring in https://danieltan.weblog.lol/2026/05/dunning-kruger-and-the-... , the expert pays the "communication tax" to dumb down concepts that the listener can understand. There is a gap between domain understanding and what is being conveyed that is similar for human-llm interactions as well.

Really it was not the issue that Opus could not do all these, there was just no incentive to fix bugs. Mythos represented a real marketing use case, so yes thanks for spending money to fix this, but this is not sustainable.

Really the problem boils down to:

“I don’t want to maintain a custom fork with my fix” - valid sure, if you are not sure if your fix is the best solution for it, and would like the general community to comment (i think the problem here is that iterating on a fork is generally difficult to discover and work on)

“I really want everyone to have the benefit of this fix” - Could be interpreted as wanting the "fame" of authorship or participation in a large open source repo, otherwise just sharing the fix and letting whoever needs it is enough tbrh.

“I don’t know what the fix is, but there is a bug here and the core team should fix it” - would be a user support issue.

I want to make my stance clear.

Bell's theorem rules out local hidden variables, not hidden variables. An ontology-first approach to QM (rare in modern physics) takes Bell as the constraint it has to operate inside (same as Bohm-de Broglie) rather than as a settled debate.

The Bohr-Einstein debate wasn't adjudicated on substance. Copenhagen won sociologically, and the renewed traffic on it is the field acknowledging that "settled" was always a cultural claim.

My analysis is on diagnosing the methodological inversion that produced the impasse, rather than picking sides inside it.

Sigh. I meant that the zig authors did not make it a general pattern and just slapped on the DI pattern specifically for io, instead of generalising the abstraction so people can DI stuff.

A common workflow is to run code to test some function in the REPL and then promote it to a test when you are ready, and this process has been the smoothest in lisps, especially since you can create your own test harness if you need to.

Fun fact is that giving AI repls also reduce error rates so much that you can save up to half the tokens/time or more.

I am not even sure if its a general pattern (inject any dependency?) or a specific pattern they added to Zig

Which is why Clojure advocates for simple data structures. They compose easier than code itself.