Raw LLM Responses

My problem with what you're saying is that you're advertising "no hype", but then you go on to measure productivity in terms of code output, talk about ridiculous timelines in tech, talk as if the "skills" of using "agentic" LLM's are actually at all difficult for technical people to pick up quickly. I'll take each of these points in turn: 1. Code output is notoriously a horrendous metric for the health and efficiency of a codebase. 2. You say, "If you're not someone like that and you spend even, you know, a month disconnected from your computer, looking at the news sites for the next AI model, you're just already behind." But if we look at the openclaw (or clawd bot or molt bot or whatever stupid thing it is now) what would change if you took that month away. Literally nothing happened except a slop product ensloppified and now no one cares about that slop product anymore except the niche. That's fear mongering. That's not "No hype" as you say. 3. It takes very little effort to start using an LLM harness. There's a million of them. There are a million techniques from skills to MCP to whatever that all exist in the harness. If you give a technical person access to it and they're motivated to learn it, the "skill gap" isn't that wide. It will take a maximum of 1 week for someone who's legitimately cracked at SWE as an old head to get really good with "agentic" tools. This also seems like fear mongering. Here's my experience: They produce a lot of trash if you're not guiding them extensively. Not a little trash. A lot of utter nonsense. If you're not guiding the models in a very particular fashion, a fashion that you know works for your work flow, it will produce slop. For crissake, it's bizarre that anyone puts that much stock in something that still to this day the underlying models can't count characters. It is impressive how many problems can be boiled down to next token prediction, I agree. In fact, it surprised everyone how many problems it can solve. But let's not get it twisted. You are an influencer. You benefit from the hype. Coding is completely tangential to you at this stage in your career. No one should assume a lack of bias in your take especially when you have clearly demonstrated the opposite.