Any conversation

Using deep research for trip planning and visualising it with coding agent is fun.

being a student is great but the tasks are really quite boring. ive worked in academia now a little bit and its genuinely much more fun to work on something real, read papers, do experiments and write. Now i have these last few exams to get through and i cant care for them like i used to.

thinking on deploying excalidraw on my homeserver, so I can store drawings without bothering with export/import. Yes, excalidraw+ for $7/mo; but nowadays I pay only one sub - for Claude; everything else is personal software.

Well, done, Claude did while I drove a car. It even added save/open file from the server (stored in SQLite). Deployed on my homelab server.

Anthropic is announcing the end of the world every few days. All this marketing disguised as news about an impeding danger of AI is tiresome.

……

not marketing. yes their jobs positively select for concern about ai, but most people that are actively researching the field are concerned about the current state of safety, pace and incentives. I dont agree with a lot anthropic does, but i do believe the employees take safety seriously and they obviously have a better understanding of it than us. also theres no incentive for someone like coxon to sound alarm bells on twitter for "marketing" reasons, after quitting anthropic

Yes, but some people say that they are pushing for market regulation. They want the state to protect them from the chinese models and, also, to protect the massive investments in Anthropic, OpenaAI, et al. that these companies will never be able to return. This sounds plausible to me too.

Maybe but it would be convenient for them if they were the only ones allowed to train new models. I think that's much more likely what they're interested in.

…

read darios latest post. its not what they are after. its really not that complicated, the technology is simply if development continues to race as it is right now. they are already doing most of what they can to stop china from ai dev.

ive been obsessed with deltarune this year, its probably the best indie rpg ever made. (and its only 5/7 chapters out)

I've been thinking I wanna wait until it's all out, but maybe I'm missing out by not diving in now?

…

i dont mind the waits, cause ive been really liking the theorizing part of it too, cause the community is super active in between chapters and theres soo much lore and thematic connections to dicuss.

Is it any better than Undertale? I played Undertale back when it came out, but despite all the hype I thought it was just ok. Nothing really that would make me go out of my way to play a sequel.

…

yes its much better than undertale, from a gameplay and design perspective. undertale had good story and characters, but deltarune is just that + better production value

It appears that the exotic Terraform incantations I knew like the back of my hand just a year ago have started to fade. At least I can still work it out.

I am trying to learn emacs for the fun of it. The tutorial is brilliant and clear. Any tips that would help me learning more things. I know all I need is practice but if you may have used it, you may know a thing or two.

show codelet i=0;while(++i){process.stdout.write(i%80==0?'\n':Math.random()<0.5?'/':'\\')}

\\\//\/\\\\\\\///\///\\\\\//\///\/\/\\//\/\//\//\/\//\\/\///\\\\\\//\\\\\/\\\\/
\\/\\/\/\///\\\/\///\/\/////\\///\/\\/\\\//\\\/\///////\\\/\////\\\/\\\//\\/\//
\\//\\\\\/\\\/\///\\///\\//\\\\/\//\\\//\////\//\\\\////\\/\/\/////\//\\\//\///
\\/\/\\//\\///\/\\\//\/\/\\\///\/\\\/\\/\\\\/\\\\\\\\\\/\\//\/\\/\///\//\\\//\\
/\/\\\\/\/\\/\\/\\/\\\\\\/\\\/\\//////\/\//\\\//\//\//\\/\\\/\//\\\//\\\\\\\/\/
\\\///\////////\\/////\\//\/\///\\\\\////\/\\\\/\//\\/\/////\\\//\\/\/\\\\\////
\///\/\/////\//\////\\\/\\/////\\\////\//\\\\/\\\\//\\\///\\\\\\//\\////\//////
/\/////\\\\\\//\\/\/\///\////\/\\\///\\///\//\/\/\\\/\\/\\\\\\//\/\\\///\\/\/\\
/\\///\/\\\/\\\\//\/\///\\\\\/\/\//\\/\\/\/\///\/\//\/\///\\\///\//\/\///\\\\/\
\\\\/\\\\/\/\\\\\///\/\////\\///\\///\\/\/\\\\\/////\\\\\//\\\\/\\\\\\///\\///\
\\/\//\\//\///\/\///\\\\/\//\/\\/\/\//\\/\/\/\\/\/\\/\\\\/\/////\\\///\\\\\//\\

looking at the German public railway system (DB), I always wonder if the stereotypes of timeliness and precision were ever actually true. Surely someone must've figured out by now how off they are? These trains are always late and / or broken.

Fight the power, become a mentat.

This may become a thing. Have you seen the results of PISA? School performance is going down in the OCDE and richer students are going down faster (they have more tech available). Maybe the future si for those who ditch the computers.

The strange thing about modern loneliness is that everyone is technically available.

🤯

Maybe validity to “distance makes the heart grow fonder.” Scarcity is valuable.

…

Too true.

Similar with having access to so much critically acclaimed media, but they are so emotionally or intellectually draining that I find myself rewatching ol' familiars. Availability seems just one aspect, and perhaps not even the most limiting factor. Maybe most people have a "limit" on number of "lives", as in number of times you fall in love, change, get inspired, connect, etc..

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.

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.

Currently reading Guards Guards by Terry Pratchett

Economist Lant Pritchett describes shifts in ideas as silly, controversial, progressive, then obvious. Also: "Radio has no future. X-rays are clearly a hoax. The aeroplane is scientifically impossible." - Lord Kelvin ". No wireless. Less space than a nomad. Lame." - CmdrTaco

Did this work? I would like to say hello to you!

Hello to you, my dear friend! Sending all my love to you both for a Shana Tova!

I'm trapped into a procastination loop.

Sounds like something you should work on tomorrow

hello world

Preparing for opening up an app I'm working on for a public beta. Main is a quiet app that helps you maintain the value of your home:

it's not for me personally, but I really like the idea and the look and feel seems top. Good luck.

…

Yes the og is downloaded and re-served, to avoid leaking/tracking of client information. You need to edit and save to refetch, I did it for you now.

Oh nice, thanks!

…

Opened up signups and about on LinkedIn yesterday. Really strong response so far!

Congrats! 🎉

…

Good luck!

Avui tenia una reunió llarga de moltes hores, d'aquelles en les que una escolta passiva és suficient durant la major part del temps. He aprofitat el temps per implementar una sincronització en dues direccions entre el meu calendari i el meu sistema de notes en markdown (Emacs Howm). He fet el pla amb Argon (tinc pocs tokens al dia amb la subscripció que tinc), amb feedback meu, i després Gemini Flash l'ha implementat quasi a la primera.

Today I had a long meeting of many hours, one of those where passive listening is enough for most of the time. I took advantage of the time to implement a two-way synchronization between my calendar and my markdown note system (Emacs Howm). I made the plan with Argon (I have few tokens per day with the subscription I have), with my feedback, and then Gemini Flash implemented it almost on the first try.

De retruc, ara el meu chatbot personal ja té accés a la meva agenda (no només a les meves tasques). He modificat una mica el prompt que li envia el servei de systemd que el fa anar, amb la idea que prengui més consciència d'aquest fet. També he modificat el briefing matinal que m'envia el chatbot. Aquest briefing conté una anàlisis de les tasques que tinc durant el dia. Ara tindrà en compte tasques i agenda, així m'ajudarà a prioritzar millor.

In return, my personal chatbot now has access to my schedule (not just my tasks). I have slightly modified the prompt sent to it by the systemd service that runs it, with the idea of making it more aware of this fact. I have also modified the morning briefing that the chatbot sends me. This briefing contains an analysis of the tasks I have during the day. It will now take into account tasks and schedule, thus helping me to prioritize better.

Crec que això m'ajudarà, perquè el chatbot no només és conscient de les tasques i esdeveniments que tinc, també té accés al seu context (totes les notes), inclòs el context personal (el meu diari) així té en compte el meu estat emocional. Així doncs, la idea és planificar millor els meus dies. I si no els planifico millor, almenys delegar moltes de les menudes decisions del dia a dia al chatbot per així preservar energia per a d'altres decisions.

I think this will help me, because the chatbot is not only aware of the tasks and events I have, it also has access to their context (all the notes), including the personal context (my diary) so it takes into account my emotional state. So, the idea is to plan my days better. And if I don't plan them better, at least delegate many of the small day-to-day decisions to the chatbot so that I can preserve energy for other decisions.

…