pneumic

The sunny side of the road?

Opus has been fine on proteomics and bioinformatics for me. I have never seen a Claude model refuse on such grounds before in the past.

Claude is still the best IMO, but it feels like its most frustrating and grating aspects are not down to the model’s abilities, but the increasingly heavy hand of Anthropic expressing itself within the model. Fable’s comically useless responses almost seem like a cynical marketing tweak.

“This model is so powerful we basically can’t let it do anything. How terrifying! We need more money to make it stronger. Now do you see why we should be the ones who write the regulations? We’re the Good Guy AI Company Who Will Never Ever Ever Be Unethical after all.”

As this entity gains more ground, their models become increasingly annoying to use and their little act becomes more transparent. The whole “I’m-just a befuddled ethically-minded AI researcher who is perturbed by the power that I unwittingly discovered and I must warn the world” thing? Yeah fuck off. Your twee pandering to naïve nerds and cynical technocrats is nauseating and ordinary people can smell it a mile away. Completely repellent leadership who put up red flags to anyone left with a working ability to read between the lines of both spoken language and body language. The tech company equivalent of a sex predator who plays as the nice guy. Gross.

Nobody likes these companies and their models are annoying, but we’re going to put up with playing middle manager to these obnoxious programs because our jobs depend on it now, and these products are still the best on the market.

A breakthrough in tools that facilitate user-owned models and infrastructure is desperately needed for the sake of our dignity and sanity, if nothing else.

I work on software that talks to mass spectrometers and it consistently refuses to refactor even an input file parser, presumably because it can infer it’s related to biology? Useless indeed.

Came to post exactly this, except it’s got me using emacs again. I led myself into some mild psychosis where I attempted to mimic the Acme editor’s windowing system, but I recovered

Give them love and care, and let them grow out of it. Do not sterilize and mutilate them—a most extreme form of “conversion therapy”.

Putting money down on this being primarily LLM coded

AI will make humans more AI-like, and milestones will be celebrated when it more perfectly simulates degraded humanity

For me LLMs have been an incredible relief when it comes to software planning—quickly navigating the paralyzing quantity of choices when it comes to infrastructure, deployment, architecture and so on. Of course, this only highlights how crushingly complex it all is now, and I get a sinking feeling that instead of people solving technical complexity where it needs solving, these tools will be an abstraction layer over ever-rolling balls of mud that no one bothers to clean up anymore.

Wezterm is actually programmable. I am looking to drop Kitty as it intentionally offers minimal tmux support and the text rendering options that made it superior for me are being deprecated.

Until Ghostty offers the scriptability found in wezterm and kitty (e.g., hit a keybind, spawn a new terminal and execute a font picker script), I am trying out wezterm, which is pretty great, but renders fonts too thin by default. I stare at this thing eight hours a day so text rendering is super important.

I am lucky to have never had a live coding interview, because I would utterly crumple. Not proud to say and something I should work on, but it’s very real.

This is proof that they aren’t a European country.

The genteel American Beech is currently threatened by disease. Where I live in New England is covered in beeches, and starting last year I have not seen a single one that doesn’t show symptoms of infection: https://www.fs.usda.gov/inside-fs/delivering-mission/sustain...

100% agree, especially on the lispy expressiveness. I love that I can build analysis pipelines in a functional style, which has always clicked with me more than other paradigms.

Tidyverse is a godsend for at least getting initial data transformations sketched out and for gently introducing new users, but I do believe one should gain an understanding of how to do all of these things in plain R.

Nice post. R's quirks seem to put some people off but I've found that it's a relative joy for exploratory analysis and visualization like this, especially within RStudio.

Recently I was tasked with grouping a large number of DNA oligonucleotides, and exploring the criteria by which to group them was a lot of fun using various R libraries. In the span of a few days I learned how to use k-means clustering, how to employ an UpSet plot, and how to build a phylogenetic tree.

Aqua Voice: https://withaqua.com/

Ah yeah, I do this too. The hostname and current working dir gets its own line, then on the line below I have my prompt sigil (% or $, depending on shell). What I meant was in the new Jetbrains IDE terminal, the prompt is pinned in a pane at the bottom of the window while you view your output doing its thing in the pane above.

Could you elaborate more? I'm having trouble visualizing what you mean.

At first glance, I love the idea of the prompt being in a static position, separated from the output. Surprised that this is the first time I've seen something like that.

All the time:

"Text $WIFE_NAME I'm on my way home"

"Directions to $DESTINATION"

"[Set a] Timer [for] 15 minutes"

"Call $CONTACT" or "Call $PLACE"

etc.

These are usually done while I'm driving or on the move, or otherwise don't feel like scrunching up my thumbs to type or to look at a screen

I couldn't bear to finish reading your story, for which I apologize. Our son is physically and developmentally disabled and though not at all near the level of yours, he will probably require assistance for the rest of his life. I will think about your family and story often, and may God keep you all, especially your beautiful little boy.

Where are there so many seemingly dishonest Alzheimers researchers lately?

Seeking advice on how to lobby my city to disable a streetlight. We are on a rural road next to conservation land, and it blasts our house with light through the night, distracting our son with autism. Considering getting an instrument to measure the light's intensity and gathering data. Any tips on how to lobby a city government with such an unusual request would be greatly appreciated.

I planted some in my yard up in New England, where it’s native. It’s spread nicely, and seems to act as a “sink” for aphids, keeping them away from other plants. They flock to it but don’t appear to seriously damage the plant’s ability to do its thing.

This analogy is naïve. There is plenty of evidence of extermination behavior among early native Americans. In more recent times, the Anasazi were especially feared by their neighbors for the horrors they'd inflict.

Neither have I. My main complaint is that it's slow.

Yeah, it's not bad news at all but it appears to be one of many that still require considerable heat, industrial infra, etc.

Has anyone attempted to put these enzymes to the test in at some scale in an industrial setting, I wonder?

on mac I do this:

    alias ls='LC_COLLATE=C \ls -F'

doesn't sort dot directories before dot files, but it's good enough

Prigogine/Prigozhin is a not-uncommon Russian-Jewish name: https://en.wikipedia.org/wiki/Prigozhin

My parents were mild hoarders, but it was enough to cause me anxiety growing up. Incidentally, after patching up their marriage, healing their relationship and addressing long-standing problems between them, they finally began to clear it away.

Looks like 9term, the Plan 9 terminal emulator.