Yes. The models may have started from indiscriminate scraping, but people are undoubtedly working on refining the training data. Combined with the overall model capabilities, I suspect code quality will continue to go up.
What you're suggesting is a negative flywheel where quality spirals down, but I'm hoping it becomes a positive loop and the quality floor goes up. We had plenty of slop before LLMs, and not all LLM output is slop. Time will tell, but I think LLMs will continue to improve their coding abilities and push overall quality higher.
I find agents work well with Elixir, but you should reach for it when the product benefits from Elixir/BEAM features. A slow compile time is a minor annoyance compared to the greater architectural decision. Elixir hot-reloading with Tidewave works well for agent loops though.
> Does Claude leverage the trailers automatically, or is usage initiated by you?
Trailers hint is in my global CLAUDE.md so it knows: when debugging, saying something like "didn't we already discuss this in a previous session?" it will know what to look for.
I also have a manually invoked `/search-session-transcripts` that I can use to natural-language inspect previous session by day, project, session id etc. Claude often uses this skill to narrow down on parts of the conversation that are relevant to the current query.
> How often are you using the trailer lookups?
Mondays are usually the day I need to refer to previous sessions from the week before. Trailer lookups are also good for continuing buildout of adjacent features. They've also been excellent in incident post-mortems where the PR text and commit message aren't enough to gauge the "how" of decisions that led to issues.
> Any idea how this relates to token usage?
I tested this. Session-transcripts are append-only so `/clear` and `/compact` don't clear out old messages, they stay stable and accessible. I also don't clean out my `~/.claude/sessions` ever so there's a lot in there, but the info is valuable and cheap.
Still somewhat optimistic here. It's easy to fixate on the negatives, and tech has brought many negatives for sure. It sometimes feels like governments and big business read all the dystopian sci-fi novels and turned those ideas into operational playbooks. But tech has brought many good things as well.
If you're a basic bitch consumer and complainer, that's on you. The world is hard, life is hard, humans are beautiful and terrible. These things don't change. But there are more options now, and it's never been easier to learn, explore, and DIY.
Besides all the nice luxuries like power tools and google maps and online shopping, my true optimism lies in digital data, transparency, and accountability. The powers that be of course are allergic to latter two, but tools are there now and we're slowly making inroads. There are real problems being solved across all industries that wouldn't be possible without technology.
Nostalgia is fine and all, and the 90s were certainly a more naive time of course, but it's not like everything was great and now it's terrible.
LLMs are the future because you have an amazing amount of information available with low friction, plus the ability to reason (sort of) about things. In some cases they might regurgitate, but they're also pretty good at synthesizing and comparing. None of this is perfect, but nothing else is either.
LLMs are a powerful tool like we've never had before. You don't expect a chainsaw to cut down a tree by itself and carve the wood into a statue or a new compiler. LLMs aren't mind-reading autonomous creators, they're more like a mech suit that can increase your capabilities. They have flaws, but until something better comes along, it sure seems like they're the future.
PII leaks are normalized now. Most people aren't even aware, or just shrug "oh well" and head to the app store to download the latest gacha game or whatever.
Claude self-reflects and updates based on feedback pretty well these days, but seems to lean on memory more than updating CLAUDE.md. I don't know how well it adheres to memory, but it seems to work sometimes. I don't like how the memory is stored outside of the project directory though.
Hmm I would hope that's for better quality (if there's somehow model-specific optimizations) or search/retrieval methods down the line. But can't help but feel like the labs/providers might try to lock-in customers by making things non-portable/opaque.
It's cool that they did some measurements, but unfortunately there's not much to learn from the article unless you're using really outdated files that you wrote by hand. The agent should know how to write a good file.
For existing files, the agent will carry on a bad structure unless you specifically ask it to refactor and think about what's actually helpful.
In general, it should be a lean file that tells the agent how to work with the project (short description, table of commands, index of key docs, supporting infra, handful of high-level rules and conventions that apply to everything). Occasionally ask the agent to review and optimize the file, particularly after model upgrades.
Everytime I've asked a model to write it's Agents/Claude file it's been pretty bad actually, are you sure writing these files is actually in distribution right now?
I don't have a ton of experience with this, but every attempt I've made to quickly get an LLM to one-shot an AGENTS file has been too verbose in all the wrong areas. I'm not convinced LLMs are actually good at summarizing anything complex. Maybe some "blessed" prompts will bubble up in time that change my mind.
LLMs don't one-shot anything very well IMHO, but if you make several passes and work through each section it should end up ok. The key word I use is "optimize", but I also press it on what's actually effective. The goal is a small file, so just ruthlessly cut anything that doesn't have high value across the entire project.
Again, the goal is to let the agent know how to work with the project at a high level, not much else. Skills and docs cover the rest.
What you're suggesting is a negative flywheel where quality spirals down, but I'm hoping it becomes a positive loop and the quality floor goes up. We had plenty of slop before LLMs, and not all LLM output is slop. Time will tell, but I think LLMs will continue to improve their coding abilities and push overall quality higher.
reply