I'm using my ADHD hyper focus skills while flagging issues that the LLM is doing.
Reading 10x more code than before puts me straight into the zone. (In a language that I find interesting: Elixir)
My own process is improving so much that I had only one bug last week that was fixed immediately after the error tracker caught.
But yeah I feel more tired sooner. So it's oneto three hyper focus zones per day, just like before.
The difference is that I enter faster, and now I'm not afraid of leaving the task and resuming later, since I can just ask for a summary of what we did so far.
I'm using two different models from two different providers to cross check the work tho.
I'm very good with bad smells I guess, after years supervising less experienced developers since early days in my long career.
Well I do intend to use software to automate these interactions, because in my country whatsapp groups are unmanageable without this.
Imagine the group of parents of my kids schools sending 100 to 300 messages per day with different subjects.
The issue is. I also have personal and important chats that I don't want to share with an vibe coded AI software without any canaries taking the shot first.
And I'm talking as a person that is using almost all my Claude max subscription every week.
But I do verify ALL of the code that I'm delivering. And I'm even using Gemini as an adversarial LLM to review Claude generated code.
Does this that gigantic project set any standards for this?
I was not able to find on their documentation.
So it's funny indeed, but for now I'm upvoting this one even being a confident moderate person.
I use mise to update binaries. Especially TUIs that are not on the arch repos. It supports several backends, from cargo crates to GitHub releases, to uv for python and so on.
So one doesn't really need homebrew that has Linux as third class citizen (with the 2nd class empty)
Definitely. The things that I most agreed with my doctor and my psychologist was around rebuilding my daily routines on things that are important, but not overly rewarding, or with delayed reward (physical exercises, home chores, reading and entire book chapter before stopping).
So much so that I can reduce the medication when I feel my routine is healthy enough.
No overdoing neither under doing. Taking care of hyper focus is more important than anything else. Learning to transfer the rewards from highly rewarding things (videogames) to the chores.
During my gaming sessions as well, playing harder modes repeatedly without giving up.
It's very important this aspect of reframing things. Pursuing healthier outcomes by changing the routine.
Unfortunately a really close person to me is taking much higher dosages, but it's not seeing this way, neither the doctor of this person frames the same way like my doctor is doing.
I went all up to 50mg, using generics or whatnot, extended release or regular ones, now I'm a happy camper on 10mg. Will go to 20mg because I need to force myself into more exercises, usually I take 30 minutes before these sessions of "boring" routines.
More important than anything is having an honest discussion with my psychiatrist about the role and objectives of taking the medication.
I did invest my time on more specific technologies, not following the mainstream, I accepted a 35% cut, but on smaller and promising small company that smells like startup, but it's actually an small business sharply focused on some types of customers. No VC funding.
I'm also a father of a daughter and I'm "training" her to understand that she needs to sell something that is not abstract.
As a computer scientist, I still sell implementations in form of code, but I dearly assumed many years ago that this (abstract product X concrete product) could be the difference to put food on my table - and now on my daughter's table - or not.
I'm far from highly successful because I'm still fighting this urge to just focus on code, but luckily that understanding above make me talk to any kind of person that is not a coder as well. Understanding business in general, meet concrete needs on different types of industries and so on.
I'm not from the US, so take that into account. I was near shoring to the US, now I'm doing that, but to the EU. I moved to a different market.
Also I'm fully invested in using AI a lot. With that I feel as a senior dev that indeed I don't need as many juniors as before, and I dearly miss having some junior dev to ask the "dumb" questions that I forgot to ask myself sometimes. AI still does not fill that role.
What helped me with AI is not caring as much about the number of LOC as before, as long I met the product's needs.
I hope you can find some comfort, keep your temper cool, so your daughter can feel that as well. That she has at least someone to hold and float.
My best wishes and I hope that 2026 could bring better luck.
As an introvert living in Rio de Janeiro, I can tell you that a lot of being happier in a hot climate with a lot of people around is just a social mask.
When I start deep questions about financial safety, the future and so on, just by asking I can be labelled as a pessimist. And I'm far from that.
I'm a fairly resolved and confident introvert, but I know many timid people that feel ashamed that they don't feel "happy" in these large group of people, that are extremely agitated and yelling around to grab some piece of attention they need.
And what is being shown in social media, documentaries and etc is just one pov.
Is there a way (like a CSS rule or something similar) that when you look at the main strudel window, it only shows the piano rolls, punch cards, sliders, etc - but not the code?
Maybe with just the comments? This would be killer, since I have dual displays, and on one I can just focus on the code, the other one can have all the visual stuff.
I'm using this plugin, but having the code twice distracts me a lot (but I prefer the original neovim instead the integrated vim mode inside strudel).
Intelligence in my book includes error correction. Questioning possible mistakes is part of wisdom.
So the understanding that AI and HI are different entities altogether with only a subset of communication protocols between them will become more and more obvious, like some comments here are already implicitly telling.
I think it's quite the opposite, diacritics teach me how to speak a language that I'm not native.
Portuguese don't use as many as Romanian, but they are very useful. After all what's the difference between avó and avô and just avo?
We don't use in very informal settings, like Whatsapp chat, because a native reader can infer from the context. And that's how English without them works, right?
Actually I wanted the English language to have some, so the tiny differences in certain constructions would be more obvious.
My doctor let me change my dosage whenever I feel. She trusts me for this. She is also an ADHDer herself.
I've found that depending on circumstances I can do well with 10~20mg of the cheapest generic methilfenidate, non LR/XR/whatever, so in my country is USD 10/20 per month.
I went as high as 50mg of the USD 100/mo famous ones.
Over time with my other therapist (psych) I trained myself to have some discipline processing my feelings, etc. Understanding the routines that were lacking.
Here it's common to have regular meetings the psychiatrist/psychologist combo. So different perspectives.
My biggest issues were knowing what to do, but not getting it consistently, like:
- getting x minutes of sunlight during the morning, and be consistent almost everyday
- drink water even on hyper focus moments
- pay more attention to breathing even when I'm in the zone doing a lot of apparently rewarding tasks
- trying to stop on unproductive hyper focus moments, realizing when they come
But the medication is necessary, since changing habits, specially the bad ones is harder.
I use the meds as an opportunity to understand myself and having easier time relearning my habits, and getting rid of the bad ones.
I'm a heavy AdBlock user, I pay for YT premium, and I paid Nebula for 2 years, also I try to buy some albums on Bandcamp even with YT music subscription. What more they do want?
And I do use referral codes for the content creators I do like. My Amazon referrals do still work.
As a mostly software backend dev I even visualize the JS guy saying "it's solved" when he forgets to tell that the correct choice is to do the counting on the backend, period. Not hacking a crappy JS snippet calling a different host.
I obviously ask for more time to make sure it's reliable.
I literally saw something similar happening around some years ago in a adjacent team I was working.
I want to pay with money, not attention. Both at the same time? Non negotiable.
Yes, but you can have a similar setup to what he is describing, just with different commands.
I'm using niri instead hyperland.
I can either use `sed` on it's configuration file (on/off, resolution, position) or for some of its configurations I can use the cli (for output scale).
I was a happy Max user.