ergl

> The people willing to spend their precious time on them are probably the ones whose opinions matter the least.

Yeah, the people who disagree with you are never worth listening to. (/s)

> But the two ATMs problem would be solved by securing a lock on the database before proceeding with the transaction?

You can also let an overdraft occur. Now the user ows the bank some money and you can charge them interest for their debt.

> Literally how did you come to the conclusion this was written by ChatGPT?

You tell me:

> Scroll to follow one closed strait into fuel pumps, harvests, bond markets and my own country’s winter heating

Or:

> France: a run on the pumps. [...] drivers fleeing record prices elsewhere emptied its tanks faster than trucks could refill them. [...] A price cap meant as a cushion, in a system with no slack, turned a price shock into empty pumps.

A cursory check on Pangram returns 100% AI written just from the first 4 paragraphs.

Once again the top submission of HN is an LLM-generated article. I know we're not supposed to speculate on people reading past the title, but I incur you to read 3 paragraphs and tell me with a straight face that you did not eyeroll at every short, empty phrase stated with enormous confidence.

> the only real software is blessed by our sacred priesthood of PLT experts.

Incredible how anti-intellectualism keeps taking over the software engineering profession.

All data labeling is training, but not all training is data labeling. In this sense, the headline is correct, I don’t get the nitpicking.

> they fired contractors for using AI to label data

Is your objection to the word “train” in headline? Otherwise, you’re just restating the headline while calling it clickbait (which, btw, it’s not, perhaps you meant to say it is misleading, which is a different thing.)

The paper is about temporary workers, specifically for seasonal jobs. They’re not buying houses or renting long-term. In fact some of these jobs include on-site accommodation for workers.

> I'd love to see someone try to run one that can meet 100% of the world's demand for remote works.

I, too, have infinite contempt for my fellow working class members.

> Is anthropomorphizing a real problem?

The paper argues that pretending that the so-called thinking traces represent real reasoning can lead users into trusting wrong answers, if the thinking traces appear convincing enough. Researchers might inspect these traces to try to determine the “intent” of a model, as well.

For an example of the latter, when OpenAI spoke about the hacking of HuggingFace at Black Hat, they repeatedly showed the thinking traces of their model as “proof” of what the model was “thinking” as it performed the attack, calling out “surprise” moments, etc.

Now, it’s possible that the employees presenting didn’t truly believe that the thinking traces would give them useful clues, and presented them only for a “wow” factor, but I wouldn’t discount the possibility that even the people working at frontier companies can fall for this tendency to anthropomorphize LLMs.

On my knees begging that the developers of AI detectors use them on their own writing and that at least pretended to mask the obvious LLM cliches in their blog posts.

You're right, understanding things is so 2025.