Common side-effect of letting Claude write your landing page.

Apparently it's mostly this:

    // ENCODE — pack(html) → one .hmml (a Uint8Array you store or send)
    //   · lifts every data: image out of the HTML into raw bytes (no base64)
    //   · gzips the HTML/CSS/JS and frames it all as one binary blob

Well, define "drone warfare"- the CIA and the Pentagon has been operating Predators and friends for a long while.

those predators and friends are really high-altitude drones, and for them these low-altitude (human) level pics don't give them any advantage

Very likely. I'm just saying, people in the CIA seeing where the tech might be going and hedging their bets is not that unlikely.

The idea is to expose it as a tool to your LLM agent so it can run calculations on its own initiative.

If something's a bad tool that misleads people into doing bad work, it would be good to know that.

> A simple distributional analysis of every rsync release with bug data. No model. No assumptions. Just placement.

If you want me to read your analysis, you are going to have to make it not read like Claude wrote it. What does "placement" even mean here?

Yeah, made me chuckle that an LLM— probably Claude— was used to write this.

The use of "regime shift" is what gave it away for me. I've never seen a human write that, but Claude does from time to time.

At least they removed occurrences of "load-bearing".

"quietly" seems to be the new one recently

Ohhh, quietly load-bearing is the real just. No noise. Pure fact. Delivered robustly.

It's the ultimate product for marketers. It inserts itself as an advertisement into every conversation now and defends itself against criticism. Just crazy. There's no hope for the rest of us.

It's not defending itself here, both because I used GLM 5.1, not Claude, and because I was the one who decided to do this analysis, iterated through six or seven different methodologies to try to find the one that was most honest with the data that I had (all of the methodologies showed directionally and often in magnitude the exact same thing, but I wanted to do something that fit the purpose, in consultation with my wife, who, as I've mentioned elsewhere, has a master's degree in statistics), and, of course, I specifically chose all of the metrics and sources for the data.

If you don't want to read the LLM prose, you can just go to the GitHub of my project, grab the scripts, and run the full pipeline. It will gather the data, build the database, and run the analysis from scratch for you, and you can look at the numbers directly. It's all repeatable.

Your rewritten post is far easier for me to read now, fwiw.

LLM output has conditioned in me a near reflex response to just close a tab as soon as I smell LLM-authored text. Like, I'm not mad or anything, I just frequently find most default LLM-voiced text very unpleasant to read so I just don't continue reading.

"Placement" as in where the Claude-driven releases exist within the existing distribution of bugs per 100 commits. If they're not OOD, then nothing is unusual.

Also, it wasn't written by Claude FWIW, GLM 5.1.

it can operate at the level of a mere mathematics professor, who everyone knows are barely conscious, basically automatons. wake me up when it's Einstein

Pangram flags this as 100% LLM output fwiw

I am beyond tired of reading "This isn't X. It's Y"

that happens at the end of nearly every paragraph here

It's that companies like copilot/cursor are in real trouble if they are in the business of reselling expensive Anthropic tokens

But isn't the current understanding that harness is equally important as model once you get above a certain threshold, so there seems to be room to add value there.

Cursor is potentially about to be acquired by X.ai (i.e. SpaceX), unless this is just some IPO game being played by Musk. They are certainly not just a token reseller since they have their own models in addition to their own vector database approach for code matching.

> “rename this function and update callers”

I'm old enough to remember when IDEs could do this without needing a couple gigabytes of matrices to do it

(LLMs are great for anything even slightly more complicated ofc)

The first time I was impressed by AI coding was when I pointed it at some switch case monster code and told it to replace it with a strategy pattern.

And it did just fine.

So no matter what you think about vibe coding, using AI for these slightly more complicated use cases is genuinely useful.

Pangram agrees it's 100% LLM written.

But I'm sure you understand that AI-driven "AI detectors" have an error rate of probably 90%, right?

I've heard of people taking old writings from 30 years ago, feeding it to an AI detector, and being told "this is AI"

Pangram seems to be tuned to avoid false positives most of the time.