Em Adespoton

Computer monitoring software should be designed and set up to monitor hardware and software, not people.

This isn’t to say that it can’t monitor people, but that shouldn’t be how it is architected.

Identity monitoring software on the other hand… that should be monitoring people, with the goal to protect their PII from unauthorized access.

If the model is just a model, a pre-trained probability seive, what difference does it make if someone is abusive? If it causes a token spike, the user has to pay for that. The abuse shouldn’t survive beyond the context window and the model isn’t wired up to “feel” or suffer.

So is this a publicity stunt (“Our model is as intelligent and human as a real person”), or is it about social influence/normalization—they worry that if people become accustomed to abusing the prompt, that will trickle over into other social interactions?

Here’s a question: when Google blurs things, do they go back and blur historical footage too? Is there even historical street view? I thought there was.

Funny; I haven’t seen any change in uBlock origin on Firefox.

It has indeed occurred to me; none of those signs by itself is enough to flag the output as LLM generated. But when you put them all together, and combine it with the fact that the article itself claims everything else (the site pages, the anti-bot analysis and techniques, etc) was all generated by Claude, there’s little room left for doubt.

I think you’re still missing something then; the way words are bolded, the way bulleted lists are placed, sentence fragment structure, using actual <h#> structure and using it correctly, the even placement of generated images; it all screams “AI”.

One of the reasons AI uses that type of structure is that it is easy to generate nested context that way in markdown, and then convert directly to HTML, and AI agents talk markdown. In fact, pretty much every long form article I’ve seen generated by AI looks exactly like this, right down to the lack of information density.

There were definitely some places that were human tweaked, but most of this was generated in whole cloth off of the logs the website generated.

I tried to read that. Then I tried skimming for the actual content. I found some, but eventually gave up. Claude can definitely tell a long-winded story.

What ever happened to Red Flag Linux?

I suspect the headline is incorrect though; likely numerous people knew and were smart enough not to say anything.

I’m looking forward to my next round of technical interviews. The people who depend on AI usually end up looking very foolish. Especially on the questions designed to make various LLMs hallucinate.

Something I’ve been noticing recently is that while the cost per token on specific models hasn’t gone up, the provided interfaces for using those models are starting to chew up significantly larger numbers of tokens for the same tasks that used fewer tokens with older versions of the interface software just a few months ago. Likely the interfaces are applying more expensive guardrail prompts and charging the end user for those tokens — but the end result is that it costs 4x as much to get the same work done.

Yup. Entry level wages have stagnated while food and housing prices have skyrocketed.

Exact same thing here.

If you ignore any company related to “cloud” or “AI”, especially if you focus on tech jobs at companies outside the software industry, there’s still plenty of hiring fresh coders going on.

$165,000 tech jobs are still out there. Usually they require at least 10 years experience, or a masters in mathematics or data science.

Fresh out of school? Try a $48-64k job and get some experience.

Still true a decade later.