For OpenAI and Anthropic, the $100 subscriptions cost 5x the $20 subscriptions and give you 5x the tokens. And the $200 subscriptions are 10x the cost for 20x the tokens. (Tokens cost 50% as much.)

It's particularly frustrating that it's simultaneously superhuman and incredibly unreliable when it comes to cognition.

E.g. it can whip out 100,000 lines of code in the time you'd write 500...but 5% of that code will be janky, or subtly buggy, etc.

It's the "oh...you're absolutely right!" effect. I can't think of any other tools that are so powerful yet so finnicky and almost guaranteed to fail in ways that are so difficult to detect.

(A model that can run reasonably well in a ~24GB MacBook.)

Putting readers through this exercise disrespects their time. Even if as a writer you did the work of researching, reasoning, and fact-checking, you shoot yourself in the foot by running it through an LLM because there's no way for the reader to know which thoughts/research are from you. It demolishes the Ethos of the writing; readers feel they must do quality assurance on the reasoning, research, and facts.

What's going to happen long term when this style of writing is so ubiquitous that people get comfortable reading it? Is it going to fundamentally change English? Are people going to start talking like Claude?

I just want to understand more.

Also, would you be willing to share the actual text of it that you put in AGENTS.md?

Any more detail you can share? Do the others feel more "human"? Are there any that are particularly digestible/human-friendly?

I've been wondering for a while if this is just Claude because I mostly use Claude, so this is very telling.

Does anyone have a read on if this is primarily a Claude issue, or if all LLMs do this?

I have a personal theory: LLMs are *fundamentally* handicapped at perceiving what's going on in the mind of the human (this can't be "innovated away") and that's at the root of what makes them suck at conversation.

Next time you're chatting with someone, notice how much understanding is shared without anything being said. E.g. the other person might share something deeply disappointing, and they can tell without you even saying anything whether you get what they're going through. This unspoken-yet-communicated information guides the conversation. Or as another example: humans can read the room -- you walk into a room and immediately adjust your demeanor based on what you see and sense.

LLMs are totally blind to things like this, and this adds an inescapable awkwardness to interacting with them. I don't believe they'll ever grow out of this. Which thankfully implies more long term demand for humans instead of robots. :)

1. LLM thinking 2. RLHF 3. The latest frontier models

that does anything to change this fundamental "suggestibility" of LLMs.

But who knows, maybe I'm wrong.

The reason "the agent suddenly started suggesting all kinds of things to make its code more robust" is because you said you "want to build reliable software".

It's not a signal of good judgment or understanding. It's just how LLM attention works.

Can you elaborate on what you mean by this?

Given the context of this forum, I'd be interested to hear more about what's so interesting about the technology!

Sorry, you teed it up too well. I had to!

And to your point: instructing a (non-deterministic) LLM declaratively ("get me to this end state") compounds the likelihood of going off the rails.

https://www.crypto101.io

The neural connections (or lack of them) have longer term comprehension-building implications.