sieve

> Most interfaces we use on a day to day basis are the opposite, they're designed to be accessible

I read this and then I think of glass doors with handles on both sides that I don't open till I read the instruction on the affixed plate: PUSH/PULL. The door forces this decision on every single conscientious user. These things are everywhere, including software interfaces.

All great points.

We installed a 5 kWp plant last month. Grid-tied, of course. Our daily usage is in the 12-25 kWh range depending on the season and the plant produces ~12 kWh on the worst days (~35 on the best, so far).

We have been doing our cooking almost exclusively on induction for the last 6-7 years, and I have been trying to get others to switch as well. However, it is difficult to convince people of the benefits. Some Indian cooking requires exposing the food directly to the flames, so piped/bottled gas cannot be completely eliminated.

This continued even in 2013. Had to finally pay for an inverter because PCs do not have built in batteries like laptops do and work was being impacted.

Useless industries like RO water, gensets, inverters etc are so much larger in India because of governmental abdication of responsibility due to political compulsions. Once you have your own house, or live in a gated community/highrise with its own water/power infra, you get almost western levels of comfort without the messiness of the outside world.

Crowdfunding takes the guesswork out of these things, and is a throwback to the old days of patron/subscriber supported ventures. A Kickstarter will let you know exacty how much demand there is for this before you order a single piece of hardware.

I have been programming for 25+ years and the last year has been the most fun because every single useless-for-everyone-but-me idea ... I could spend a few hours on and get it working. Did 30-40 different small projects this year.[1] Mostly used by me (and some by a couple of friends).

I write software because I am interested in the final outcomes, not because I enjoy the journey, which is often infuriating because of the mistakes you make, or the crap you depend on to get your work done.

Whether people will still have jobs in three years, or thirty, only time will tell, but I feel people are kidding themselves if they think LLMs won't have an impact. We were fine with using machines to automate physical labor as we now balk at the same thing happening to the intellectual side of things.

I, personally, have done a complete volte-face as far as my views on the subject are concerned over the past year or so as I use LLMs more and more for coding and other tasks.

---

[1] There is this idea I have had of an Excel replacement: simple, purely functional, TSV-based spreadsheet with zero backward compatibility with styles as purely optional sidecar material that I have always wanted to do but lacked the time. Brainstormed a spec with Claude today. Might work on it in the near future.

> The old social contract was that the writer had done more work than they were asking of the reader, but this is torn to shreds with LLM-generated text.

The problem is that your biases kick in long before you read the stuff. And I do not see a solution for this.

I write my devlog by hand and then run it through an LLM for a quick sanity check. Grammatical errors and spelling mistakes in stuff you publish in this age of AI is lazy. But I don't do it when I comment on HN/Reddit etc because this is stream-of-consciousness stuff and using AI actively interferes with my thought process.

My code is almost entirely written by LLMs today, but no one can judge how much time I spend on it unless they actively follow my comments on WA, Reddit, HN, my devlog and other places. All they would see is a repo with commits done in the space of a day with nice multi-para commit messages.

This is because I get an LLM to rewrite the entire commit tree once I stabilize a project and have it reproduce the final state in a sequential manner with bite-sized commits. My own development process involves 3-5 word opaque commit messages and a lot of squashing/force pushes: impenetrable/cryptic for anyone other than myself.

There is also the problem with pushing people --- for whom English is a second/third language --- out of entire conversations. Last year, I was talking (on Reddit) to someone from South America who was using LLMs to write in English and a few people were so dismissive of the messages.

I see this as an unstated case of preening.

For better or for worse, the world now communicates in English. You cannot shut people out --- because LLM --- when you yourself would not take the time to read someone if they wrote in Spanish or Russian or Japanese (or the ~25 commonly used languages from my own country).

Amazon can be utterly useless sometimes.

There are booksellers on AmazonIN who list every single book under the sun, with long lead times, and who never use Amazon for delivery. They will use fake IndiaPost tracking ids (but wrong ZIP codes) to show that the book has been dispatched. You wait like an idiot for four weeks and then complain to CS who then refund your money. The legitimate seller feedback is overshadowed by repeated fake ones from the same 5-6 names.

I have been caught in this trap a few times. And no amount of complaining to CS helps.

DS is VERY talkative. Luna is less so. Still do not think, based on this little experiment, that Luna could beat DS in price: API-to-API. As part of a Plus/Pro plan? Sure.

Love it! Simple to use. Most readers are so needlessly complicated. The themes are a nice addition.

I have written about my experience. I have also mentioned the kind of code I write. It is not react/js/css heavy stuff that I see a lot of people write. So the code bases are typically in the 5-50KLOC range. Freestanding C, Python, or maybe some TypeScript. And fairly modular. I can thus run models on specific modules without having them read everything into context.

So the workflows I mention work for this kind of stuff.

Do you really think Western providers will not train on your data? I have no such illusions.

I try to keep PII out of what I share with LLMs. Otherwise, I do not see the point, really. Very little of my code is "unique." I simply approach things a bit differently. Otherwise the algorithms and code would be similar to what others with domain knowledge would write. So much of code and algorithm implementations are available in the open. And LLMs have trained on all of them.

What they most probably gain from you is your prompts and your thinking approach more than the code.

You can use the models I mentioned directly from DeepSeek, Meta and Xiaomi and not exceed $40. Were it not for GLM 5.3 blowing up a quarter of my monthly budget in 5h, we are actually looking at something like $30.

I used to read the code till around May. Now I don't. Instead I validate behavior. And have multiple LLMs verify that the code implements my handwritten spec.

MiMo 2.5/2.6, MuseSpark 1.3, DeepSeek V4/4.1 Flash and GLM 5.3 Flash are perfectly capable of following my spec and then poking holes in the implementation till there are none left.

I use Claude Sonnet and ChatGPT via the web UI. I often use Claude to come up with specs for my ideas. This is becoming less and less useful. DS4/MS13/MiMo are almost there for these use cases as well.

I dogfood everything I produce, and the models are good at collaborating with me on a spec and then turning it into code.

If Sonnet/ChatGPT suddenly became unavailable due to Anthropic/OpenAI suddenly not being able to subsidize the freemium/loss-leader experience, I probably would not miss them. Google/BraveAI already give you the AI experience during search (when you are looking for stuff to buy, or something particular). Claude/ChatGPT still have a minor edge in this use case for me right now.

I gave you the $40 option. Which is what it would cost if you used APIs on OpenRouter or elsewhere. Still beats Luna by 4-4.5x

I do not (generally) trust benchmarks. I only trust what a model does with MY code.

Forget DS. I asked MiMo 2.6 yesterday to explain ML/LLMs to me succinctly and the pointed it at Karpathy's micrograd code. It produced a C implementation called `xor_mlp`, a tiny model that learnt how `xor` worked. I then asked it to produce a model that can play tictactoe without losing (mostly). It did. It supervised the training process and produced a compiled version with multiple switches. The pi-dev session is still running, so here are actual stats

↑45k ↓35k R1.0M CH99.4% $0.019 4.2%/1.0M (auto) - (opencode-go) mimo-v2.6-flash • high

And here is Luna on the same workflow (I had to poke and prod a bit to get what I wanted):

↑141 ↓34k R1.0M W43k CH95.3% $0.072 4.2%/1.1M (auto) (opencode-go) gpt-5.6-luna • high

I expect similar results from DS41F/MS13. Closer to MiMo costs than Luna.

So the "significantly cheaper" thing may not really hold, more so when Luna has to actually read my codebase to do the stuff that I want rather than rely on world knowledge. The 8-10x cache read cost differential itself will kill the token budget.

I have used MiMo 2.5 extensively. MuseSpark and DS4 Flash are MUCH smarter than that one. But MiMo follows instructions diligently. So it has been useful as the implementer of a spec designed by Claude/Kimi.

One good thing about MiMo that I experience on OpenCode is the provider seems to cache tokens for much longer than MS13/DS4F. I have seen cache being hit for close to an hour after the last request. The corresponding timing for MS13/DS4F is in the 1-5 min range.

I am trying out MiMo 2.6 Flash as well.

MS13 is pretty sharp and has been my workhorse for the past month. It follows my coding style and commit/clean workflows referenced in AGENTS.md perfectly but has the habit of doing things without conferring with me (the Gemini problem). So you need some kind of instruction for that.

It starts failing around the 5-600K context mark, but you can have it generate a handover document and continue in the next session.

I would not use it at sticker price, but the Contributor version is priced just about right.

I am not designing rockets. Most of my work is bog standard hobbyist stuff: compilers, vms, sandboxes, system tools of various kinds, SSGs, markup languages, plain text ledgers etc. Even Gemma/Qwen running locally can manage this.

Frankly, I have no idea what people do with Opus/Fable etc. I don't think anything I do needs something that charges $50/M for output tokens.

Are they bleeding? Their multipliers seem to be reasonable. They are not offering $60 worth of usage for $10 on every model, only some. In the case of the expensive ones, it is only $15.

Given how subscription models work (not every one uses every last $ of their plan), they should achieve breakeven soon enough I guess.

My OpenCode Go stats for the last 30d:

Cached Read: ~6,500M

Input: ~150M

Output: ~20M

Approx $40 worth of usage across DeepSeek V4 Flash + MuseSpark Contributor 1.3. And a bit of both the GLM models. This is covered in a $10 subscription.

If I were to use Luna's API pricing:

$0.02 x 6,500 = $130

$0.20 x 150 = $30

$1.20 x 20 = $24

So $184. And this is assuming smaller coding sessions (<272K) beyond which Luna pricing doubles.

--

Cost wise, these models are nice for small stuff. Translations etc. Any model that does not provide multiple Mtoks of cached reads per cent is not very useful to me for coding workflows.

I guess people are slowly realizing that giving LLMs r/w access to your entire machine is an utterly insane idea.

I did a quick look-around last month and decided to start using bwrap. But manually configuring it on a per-project basis is irritating. So I rolled out something for my own use (+ a couple of friends) based on bwrap.

What I do:

- start with `bwrap --clearenv --unshare-all --die-with-parent --tmpfs / ...`

- every single file and folder and envar I need has to be mapped in. I have profiles in TOML, and `prepare/probe` commands to make this task simpler

- `--tmpfs /` means sandbox inits as `/home/user` on tmpfs unless you specify your own `home` and `user` keys.

- `--unshare-all` means there is no network inside the sandbox. So I used `socat` to run a HTTP/S proxy inside. Lets me control exactly which host+port combinations can be accessed. But this means nothing except HTTP(S) works. So no ICMP/UDP/TCP.

- profiles can be extended via extend syntax (otherwise you have to prepare/probe/manually specify everything per profile which is a nightmare). Which lets me do a base -> net -> coding chain.

Code is merely the means to an end. It does not matter who writes it as long as it does what you want. Correctly.

Stands to reason though that someone who knows programming AND has domain knowledge can get LLMs to produce much better output compared to someone who does not.

I never used to have time to make all the stuff I needed or was interested in. With LLMs, I can.

This "shipped a whole app while being in the gym" does not work for me though. It takes me a couple of days to a week to produce solid, functional software (~10KLOC). Simple tools (3-400LOC)? Yeah, those you can produce in 30-60 minutes.

I became interested in sandboxing last month after watching LLMs fail to respect basic boundaries. Well, the very expectation that they would is foolish in the first place.

I am not a fan of application-level sandboxing. The JVM tried with its security manager, and Deno with its allow/deny, but it is not general enough for me. At some point you have to assume that anything you run on your machine is possibly broken/compromised and then deal with the situation depending on your risk appetite.

This is a long story that I have written about on my blog, but I decided to go down the Bubblewrap + seccomp + socat route for the sandboxing tool I built. Let's me run harnesses and compilers and even headless Firefox in sandboxes without worrying about damage to random parts of my system.

Yes, I meant the Flash version.

I have used Kimi 2.5 and GLM 5.3 (& 5.3 Flash). Do not need them for what I do outside of spec hardening (basically, a lot of chatting).

I tend to know exactly what I want and most of the weaker models are enough to get me there. I have mainly been using MiMo, DeepSeek V4 Flash and MuseSpark Contributor over the last month or so.

I regularly hit 200-300M cached reads every day on some of the models I use. It has exceeded 7-800M on a couple of occasions. At $0.04/M, that is $8-12 per day only for cached reads.

> Basically the old theory is true - the code IS the specification.

The spec is whatever I write by hand. The code is what the LLM writes for me. The spec could be anything depending on how much detail you want.

The problem with the "code IS the spec" in the age of LLMs is that they will change stuff without telling you while hitting their immediate goal. Six months ago, I used to review every single change. Now I get the LLM to audit the code to compare against the spec. Any divergence means one of two things:

- either I have to update the spec, or

- the LLM has to update the code.

Lots of things!

- Wanted to start blogging. So built an SSG for that.

- Have an ink tank printer that must be used a few times each month or bad things happen. Have been printing Sanskrit stories instead of test pages. Was manually building booklets with typst and then using pdfimpose. Finally, decided to write an app that does MD -> booklet.

- Decided that running harnesses directly on my machine is a bad idea. Docker/Podman are too complicated for the task at hand. So I built Adamant. It started out as a basic wrapper around Bubblewrap. Then I added networking via a socat proxy. Now it can do this:[1]

  # Runs a webserver inside the sandbox that is accessible from outside
  adamant --profile minimal prepare -- ls sh fish ncdu fastfetch uname mkdir python echo
  adamant --profile minimal probe -- python 'print("Hello, World!")'
  adamant --profile minimal run --ingress 45678:45678 -- fish -c "mkdir -p /tmp/www; echo '<p>Hello, World!</p>' >> /tmp/www/  index.html; python -m http.server -b 127.0.0.1 -d /tmp/www 45678"

  # Runs headless Firefox inside the sandbox and takes a screenshot
  adamant --profile ff prepare -- ls sh uname mkdir fish firefox
  adamant --profile ff probe --timeout 60 --host bhashika.org.in -- curl https://bhashika.org.in
  adamant --profile ff probe --timeout 60 --host bhashika.org.in -- firefox --headless --no-remote --screenshot https://bhashika.org.in
  # screenshot.png is produced in $PWD
  adamant --profile ff run --host bhashika.org.in -- firefox --headless --no-remote --screenshot https://bhashika.org.in

- Building an actual harness called PonderCode for personal use that uses ideas similar to Adamant as I find TUIs irritating for the text heavy work I do. Copy-pasting is a nightmare as almost everything is space-padded.

- Building a language learning product for Indic languages. Monetization is difficult due to the tiny market plus self-imposed restrictions like "only serious learners/readers need to pay." Only time will tell if it works out.

- My Python-replacement VM+PL project is on a bit of a hiatus. Will probably revisit it in a couple of months. Had made a lot of progress in April till I decided to expand the scope and got burnt out. Never do that.

[1] https://bhashika.org.in/logs/2026/adamant-devlog-3

You have to wonder why someone did not think of this form factor sooner. I have the original X4 and have been enjoying it a lot. My Kobo mostly lies unused because I do most of my reading in bed and the X4, even with a clip-on light, is so much easier on the wrists.

Dual core Arm CPUs, Android etc used by mainstream readers are a serious waste of resources. A microcontroller paired with the SD card is more than enough. And CrossPoint is a gem! The list of pros is long and the device recommends itself. But there is one con that you must watch out for:

The screen is delicate.

I love cases but hate screen protectors, and tend to use my devices with care. So it has worked out fairly well for me the past few months. But you might have a different experience.

It does for me. You need to ensure that they are injected into context. Your models might be broken if they ignore it. As good as ignoring a prompt.