I’m an insane person who built their own coding harness, in Rust. I’ve called it Rho. Please forgive this landing page: https://rho-code.dev I just wanted an easy central spot.
Not insane at all. People should build their own instead of using other people's agents. It's easy to do and the results can be great for you. I've built several; one that is effectively "stuck" in Kakoune: youtube.com/…/playlistYouTubeLLMs stuck in Kakoune
Yeah. It's a very large project with lots of dependencies. <The system prompt is long and wastes tokens. If you have ideas for tools (or automations like the agents.sh "loop" I wrote) it's easier if you understand the harness.
Not that I think we're close to a canonical definition (as if we'll ever all agree on one) but my team has defacto adopted the definition of code that the developer responsible for it doesn't understand, approve of and essentially has now honestly reviewed. Basically anything which leads to cognitive debt.
I agree, I don't think that there will ever be one canonical definition, but I like this definition as well, as it places the onus on the actual person creating the slop. Thank you for sharing!
@fastidiousprofilefollow"Beware those that would deny you access to information, for in their heart they dream themselves your master." — Sid Meyerswrote:
I had hit a throttle now. That’s a “subtle" way to tell me to slow down.
@kev🧜profilefollowrunner. making games, music, software. getting into as much trouble as i can.wrote:
Off to handstand class. I have been known to recommend to everyone within earshot doing challenging and/or silly physical things in a community. Not bucking the trend here.
some people do absolutely improve their work output (either amount or quality) with AI, but most people just resign all their critical thinking to the black box.
if you cannot function at YOUR WORKPLACE without ai, chances are that, 99% of the time, you never was supposed to work there in the first place.
#ainotesfollow#worknotesfollow#technotesfollow
Is this true? Sometimes, AI allows for socratic thinking, this is, the conversation with the machine allows the thinker to go further as the AI thorws ideas to the conversation. Many times, the AI reasoning is not good enough, but like a friend that says a funny think that sparks an idea, the AI can help our brain to see other possible paths.
The AI doesn't need to be right to be helpful
that's only true if a) that "rubber ducky" entity is not leading the line of thinking down the wrong path, and b) the human thinker is actually doing the critical thinking, and not the machine.
The latter is where my criticism of unbridled AI use is in this case, wherein the human becomes the rubber ducky instead of the other way around.
A problem I perceive is usage of AI to launder ignorance.
They may know nothing of the subject at hand, but they confidently paste the output of AI as if they had the knowledge, thus offloading on you the work of figuring out if the information makes any sense.
In many ways it emboldened the incompetent.
I have been using
Wezterm, Aerospace on Mac, Neovim is something that I need. My config blows up on my powkiddy, (oomkilled) other than that its perfect, and on linux its vivaldib+ ublock origin, My self hosted jf + bookserver and using webdav for obsidian notes backup was the life saving trick i needed
Not exactly "recent", but recently dived into Guile Scheme and it has totally consumed almost all of my coding work. I used Common Lisp before but this dialect is without a doubt the best and simplest of the Lisp family.
Personally, mise is more "try in hurry" compatible. So it stuck.
I have tried devenv as it seems to greatly simplify the setup. However, I had to get something done in a hurry. Tried devenv (which I installed around the time they launched) and got stuck with a load of outdated nix IIRC. So I also gave a quick try with mise, got started with fnox and it was considerably easier to get going.
I'm 55 and recently self-diagnosed as "fairly ASD with some ADHD" after watching a lot of youtube videos. It's been immensely helpful (to the point of bringing me to tears sometimes) in understanding the struggles I had when I was younger. It's also been very helpful in understanding what I can do currently to support myself and work with my brain instead of against it.
I'm glad it was useful to you. This process of becoming aware of our autism is not easy. It is liberating, but it is also difficult, at the same time.
I watched a lot of videos too, but then I visited a couple psiquiatrists and a psycologist (I had other problems too), just to be sure.
What is your opinion about the ASD youtube landscape? I think that there are a few very good channels, but there is lot of noise too. I see some danger there.
Yeah. I do like them sometimes, but the good authors are careful to start the video with looots of caveats (e.g. "this is not how to diagnose, just something that's maybe helpful").
Yes. It’s helped me understand more about how things affect me and why I respond to things in certain ways, which helps me manage situations and myself better. It’s given me a vocabulary to articulate experiences that I struggled to make sense of.
People post projects/tools/repos on lobsters all the time. From the about page, "as a rule of thumb, self-promo should be less than a quarter of one's stories and comments."
I think if @stagasprofilefollowcreative technologist
creator of textlog intends to be an active user of the site henceforth, it'd it should be kosher. Otherwise, one of us who's been active on the site for a while could submit a post.
I finally bit the bullet and have started to use dictation software more often, even outside of just my prompts. I really like it. Hopefully, computer use becomes much smoother soon. I would love to be able to control my computer with just my voice.
Trying to gently work my way back to the target intensity needed for the Norwegian 4x4 to get my weekly dose of suck and feel accomplished after. Chronic left leg pain unfortunately likes reminding me it's there though.
So, I'm watching two shows right now - Silo and Attack On Titan. I'm nearing the end of S3 on Silo and have just finished S3 of AOT. The parallels between the two are interesting but the contrasts are more interesting, namely how much more intricate AOT is.
ai coding feels like it has reached diminishing returns for me. luna is my default orchestrator, routing work by complexity. in practice, usage lands around 16% luna, 48% sol, 30% astra, and 6% terra. at this point, i care less about each new frontier release. the system already finds a capable model for the job. the frontier is moving beyond my day-to-day work, so i’m happy to settle into predictable models for a while.
#ainotesfollow#programmingnotesfollow#agentnotesfollow#technotesfollow
I’ve found the same, albeit with Opus — it’s hard to find uses for Fable. I don’t need a model that works for longer or works on harder problems, I need one that consistently produces simple, easy to understand, code; which at this point none of them do.
I've been there since I settled on a combination of GLM, DeepSeek and MiMo for planning, implementation, and verification flows.
I'm a lot more interested at this point in having cheaper models that can consistently do the work I need. It's been a while that I hanven't noticed any improvement when I try new frontier models. I only notice they are very expensive.
same here, for almost an year. At this point, I could easily live with models that were released 6-12 months ago and continue doing so.
also possibly the reason why the current focus has been around agentic workflows or agentic anything, than the model capabilities.
Coding probably yeah, but still they are bad at generating novel information, new ideas and such. They’re also not very good at seeing the bigger picture, they will narrowly perform the task but they’re not taking into account future revisions, which ends up in refactor after refactor in order to move forward.
So much present tense in #ainotesfollow generated prose. Terse, authoritative sentences, without breathing room, without hesitation or qualification. It jumps off the page, and not in a good way.
What about the lack of density to paragraphs, the length of the sentences, or the confidence and determinism of their direction? Conviction with no limits. So reliably repetitive that you start seeing patterns as if you are looking and reading a record player with some scratches on it. Skips. Squeaks. Repeats.
Yet. It gets the job done.
Yet. It gets the job done.
Here's where we disagree.
Using these tools as part of the job, certainly. But if I am to read, discern meaning, analyze choices of diction and structure, consider the context both of the author and the audience, and finally derive value from the written word, then generative #ainotesfollow falls short on many levels.
If the "job" is to produce something no one was ever intending to engage with, then we have a more fundamental question about the value of the job.
Maybe what I messaged read as a disagreement and mentioning that they get the job done was actually not about producing prose (of any quality).
I wholeheartedly agreed and still agree with you.
Getting the job done was more about producing output - be it some prose, images, videos, software, whatever. They do generate output. It is not to say that the output is meaningful or meaningless. Some souls do enjoy it. Some others - don’t.
I did start with a question though.
