LLMs have functional states that correspond to those emotions. In particular, you can extract a concept vector which corresponds to a given emotion, and steering with that concept vector causes observable changes in behavior which roughly correspond to the expectation for the analogous emotion. Anthropic (and Chris Olah's team in particular) conclusively demonstrated this: https://transformer-circuits.pub/2026/emotions/index.html

Of course it knows what it output a token ago, that's the whole point of attention and the whole basis of the quadratic curse.

"Hi there, I am Loïc Baumann, I’m from Paris area, France I develop, since early 90s, first assembly, then C++ and nowadays mostly .net.

My area of interest are 3D programming, low-latency/highly-scalable/performant solutions and many other things."

Compare that style to what's in this most recent blog - mildly ungrammatical constructions typical of an ESL writer, straightforward and plain style vs breathless, feed-optimized "not x, but y", triplet/rule of three constructions, perfect native speaker grammar but an oddly hollow tone. Or look at this post from 2018: https://nockawa.github.io/microservice-or-not-microservice/ It's just radically different (at a concrete syntactic level, no emdashes). I'm sure he has technical chops and it's cool that he worked on DOTS, but I would bet a very large amount of money he wrote the bullet points describing this project and then prompted GPT 5.3 to expand them to a blog post to "save time".

- "It’s not an accident — it’s driven by the same physics." The classic "it's not x, it's y", with an em-dash thrown in for good measure

- "Typhon brings these into the component storage model — not as bolted-on workarounds, but as first-class citizens." More "not x, but y", this time with a leading clause joined by an emdash

- "Blittable, unmanaged, fixed-size, stored contiguously per type — that’s the ECS side." Short, punchy list of examples, emdash'd to a stinger, again typical of LLM writing

- "Schema in code, not SQL. Components are C# structs with attributes, not DDL statements. Natural for game developers, unfamiliar territory for database administrators. If your team thinks in SQL, this is a paradigm shift." This whole mini-paragraph is the x/y style, combined with the triplet / rule-of-three, just at the sentence scale. And then of course, the stinger at the end.

Definitive, no, but it certainly has a particular flavor that reads as LLM output to me.

"Precision, for LEGO, isn't an engineering choice, it's a brand promise." - The classic "It's not just x, it's y", just minus the "just".

"One philosophy optimizes for cost, the other for perfection." - Again we see the x/y structure; AI writing often features these forms, eg comparisons (x vs y), conversions (x into y), negated emphasis (not x, but y), etc.

"When you have multiple parts in an assembly, use statistical analysis for tolerance stack-up rather than worst-case math. Traceability matters. Track your defects so feedback turns precision into reliability." - More x/y followed by a short stinger ("Z matters"), and the closing sentence again follows the "x/y" pattern.

For funsies I tossed the whole thing into a purported AI detector and it said 90+% confidence of AI. I don't trust those types of things very much and suspect they have high false positive rates, but I have read that AI writing generally has measurably lower entropy, so maybe it's plausible, and in this case it aligns with my existing beliefs, so it obviously must be true.