I wouldn't worry about it just yet - this is all very novel, and there's a lot of excitement involved in figuring out what it can do and trying different things.
If you're still addicted to it in three months time I'd start to be concerned.
For the moment though you're building a valuable mental model of how to use it and what it can do. That's not wasted time.
My question would be how much we think the processes will change as the models do. Much advice from two years ago is no longer relevant or realistic. Where do we think it will go next?
Does anyone have a really good way to explain to their relatives and friends how using an agent is different from simply using Google? Just saying ‘fundamentally different’ doesn’t go very far; the best I’ve found is sitting down and giving a demonstration.
It’s also difficult to explain the enormous gap between frontier models and the free ones many people are accustomed to using. Is there a tangible comparison to a normie real-life ‘thing’ that anyone has used successfully?
ChatGPT and Claude can both execute code now, so a safe subset is to show people how to upload files there and have them do useful things with the data.
You’re mistaking domain expertise with tool expertise. You can’t teach a non dev how to use an LLM effectively for dev without teaching them to be an experienced dev. Once you have that knowledge, LLMs aren’t that hard to use.
1. Make artifacts. If you're doing research into a tech, or a hypothesis, then fire off subagents to explore different parts of the problem space, each reporting back into a doc. Then another agent synthesizes the docs into a conclusion/report.
2. Require citations. "Use these trusted sources. Cite trusted sources for each claim. Cite with enough context that it's clear your citations supports the claim, and refuse to cite if the citation doesn't support the claim."
3. Review. This lets you then fire off a subagent to review the synthesis. It can have its own prompt: look for confirming and disconfirming evidence, don't trust uncited claims. If you find it making conflation mistakes, figure out at what stage and why, and adjust your process to get in front of them.
4. Manage your context. LLM only has a fixed context size ("chat length") and facts & instructions at the front of that tend to be better hewn to than things at the end. Subagents are a way of managing that context to get more from a single run. Artifacts like notebooks or records of subagent output move content outside the context so you can pick up in a new session ("chat") and continue the work.
It's less fun that just having a chat with ChatGPT. I find that I get much better quality results using these techniques. Hope this helps! If you're not interested in doing this (too much like work, and you already have something that works), it's no skin off my nose. All the best!
Not that I’ll start working on each of these this year. But even prioritising them eventually.
Claude Code gives me the courage to imagine that I’ll have actual progress on big things because it helps me maintain an overview and not get stuck on details or gas out or lose interest, which I tend to.
I recently rediscovered that I had done progress on the issue tracker 8 months ago and that this was so well-made in retrospect it motivated me to pick it up and continue. Leaving something in a well- documented way with weeks of effort poured into it over days is just a gift.
You probably aren't addicted to CC, I suspect you are just hopping from idea to idea too quickly because these new tools allow for it.
I spend a lot of time at my org doing one of the following things:
1. figuring out how to onboard engineers and bootstrap them to do their own work to draw down their backlogs
2. show the team cool ideas to spark their interest or bring them from “I’m never using this useless crap AI” to “oh, wow, I never thought of that… fires up a terminal and creates own cool thing”
3. creating a backlog of other things people want to automate but never wrote down/thought through that Claude can do in short order for immediate value
i am waiting for someone to surpass the original steam engine sankey diagram: https://commons.wikimedia.org/wiki/File:JIE_Sankey_V5_Fig1.p...
the diagram is calculated with d3-sankey & rendered with imgui. you can stretch that window however you want—ok, enough.
edit: the original steam-engine sankey diagram created a far stronger emotional reaction in me. the aella chart is just more modern, i guess?
It could have taken you years to realize that "oh I'm just exploring stuff and have no output".
Set an ambitious goal that is achievable using Claude Code, and focus on delivering it. Even if it doesn't turn out to be a hit, the experience of releasing it and using AI to accelerate it, will be a talking point to your 10-year-older self.
Discussion: https://news.ycombinator.com/item?id=46934404
Jokes aside, my relationship with my wife is always top priority, kids are priority, health is priority. With those 3 I have about 2 hours per day for myself and yeah, you can use them however you like
I'm not sure if I will look back on my life and think that I know too much and didn't produce enough widgets.
This community is obsessively pro-AI. Asking here is the equivalent of asking the guy who has sat at the slot machine next to you for the past three hours if he thinks you have a gambling problem. Of course he's going to say "no" or try to justify it, to do otherwise would be to admit to himself that he has a problem.
I don't have advice for you, other than to look up what gambling, drug, or alcohol addicts do. The path to recovery for all addiction is long and painful, but it can be done. Good luck.
We're all in the same boat right now. Except for those who decry LLMs and loudly await the relevance of their artisanal coding skills to re-ascend. :)