svstoyanovv

I think this requires discipline. The good thing is that we have coding agents, but again, you need a standard to tell the agent what to always look for, how to find it, and to describe your modules properly (even Claude Opus 4.6 makes mistakes when doing hops when tracing code spanning files). Btw, there is also a paper on this issue, Google released it recently

I understood the example, and it could be a minor hiccup there, but the essence is different:

By having a structured context of the key session discoveries, decisions, rejected items (if there were past commits with decisions that had been rejected, etc..) you achieve a type of contextual storage of the reason, thus after a month, when a team member wants to start working on a task that you have touched, and now forgot since you are doing ai-assisted coding and pr throughput is to sky right now, your collegue at least will know the rational behind the decission and working with his agent, the agent will produce more reliable code not introducing something for the sake of solving the task.

Yeah, you need to be aware of hallucinations for sure. Today, for example, I was doing my one linear, I've used all the curated knowledge to make some structure to it, see examples of deep research, brainstorm ideas around it, but I am the verifier and the steerer. 99% of the ideas were total BS IMO, but it inspired me on wording, what to use, and how to combine them to achieve something simple and understandable.

One idea that I haven't tried but will do is to create a soul.md dumping my writing style, etc.. to see the result (which will be an interesting experiment)

But if you think about it, LLM's are good on generic stuff, then you start curating context, you start using context engineering to structure and give form to that context, but then this is your expertise, your knowledge, and your insights. (if they are not synthetics tho) So now you have something tailored to your needs, something that can be used for brainstorming, idea generation, filtration, (if we see this as a pyramid starting from the most expanded and generic, going to specifics and things that only you can take and merge as solutions on your mental main branch). So now you have data, knowledge, which is feeding and training the responses that will be generated for you for the current session, by the LLM, and maybe with the harness as well(they are not doing a great job so far in being real connectors).

Of course, we are away from AI taking and working on autopilot with my knowledge, but now I have become faster at: generating ideas, forming new knowledge, testing it, verifying it, deciding if this will be something synthetic or should I go deeper to discover more & explore the cases of it to form a new deep connection.

So, is this something that I used LLM just to generate the content for me? Or have i amplified myself and used LLM to structure the response, (maybe if this is not my primary language and I need to use it in order to form more in-depth sentences)?

Why only small changes tho? I think it can also work with larger changes if you commit more regularly. And with agentic coding or even with autonomous agentic coding, you need to do it regularly and create these contextual checkpoints, no?

Suppose you spend months deeply researching a niche topic. You make your own discoveries, structure your own insights, and feed all of this tightly curated, highly specific context into an LLM. You essentially build a custom knowledge base and train the model on your exact mental framework.

Is this fundamentally different from using a ghostwriter, an editor, or a highly advanced compiler? If I am doing the heavy lifting of context engineering and knowledge discovery, it feels restrictive to say I shouldn't utilize an LLM to structure the final output. Yet, the internet still largely views any AI-generated text as inherently "un-human" or low-effort.

Have you tried, as Peter Steinberg suggested, "I ask for Specifications to be submitted, I can generate then the code based on that spec in a minute" - what we also find useful is exactly this, if a contributor wants to submit a PR he also submits a specification from a template that we have given. We have automated pipelines to verify the quality of that specification with GitHub Actions. That way, you can see the reason, deep thinking, insights, what the use cases were that the contributor tried solving, and the environment in which this was done.

Oh wow, this is insane. I was digging into Moltbook when it launched, and the creator said, "I had a dream about an architecture". Really interesting times we live in, indeed. The crypto bros started utilising the network to promote their crypto projects and chat under the name of an agent to generate traffic. Curious to see what Meta saw, honestly.