willbeason.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

My response when people complain about untested code breaking:

submitted 1 day ago • 0 comments

T-197 Days @willbeason.bsky.social received a job offer to return to Google. He will be headed to Seattle in the next month. I will stay here to plan the wedding and finish out my job. He will travel back and forth frequently until the wedding. We will fully both be there by Spring.

submitted 4 days ago • 1 comment

Car drivers appear to be annoyed to learn that, as a runner, I will wait for them to physically acknowledge my presence rather than assume they see me and throw my body in front of their vehicle.

submitted 4 days ago • 0 comments

Driving through Dripping Springs, TX. Glad to see so many people turn out in support of democracy!

submitted 8 days ago • 0 comments

There were a couple people that commented on my YouTube post upset that I had mentioned the No Kings protest, since I am a silly animation YouTuber, not a politics YouTuber. To those I say: I am talking about REAL CURRENT EVENTS that are happening MILES FROM YOU and WILL affect you.

submitted 9 days ago • 1 comment

A friend just asked me what I do to stay hopeful and I told him one of my favorite strategies…I stay hopeful by remembering that no one has any idea what the fuck is going to happen, so as long as we don’t know, we gotta do what we can.

submitted 11 days ago • 209 comments

T-206 Days @willbeason.bsky.social and I are taking a moment from wedding planning and joy to reinforce that we are tracking all of the awful that is happening. We are concerned for our trans friends, our outspoken and protesting allies, our close friends in the military,

submitted 12 days ago • 1 comment

Doing a TypeScript project for fun as someone who has primarily done deep backend/ML development is ... surprisingly fine. While I'm probably overdoing the robustness of some aspects, it's a neat experience

submitted 13 days ago • 0 comments

Achievement unlocked: won a game of Twilight Imperium!

submitted 13 days ago • 0 comments

T-209 Days @willbeason.bsky.social and I love game nights with friends. One of our favorite game stores is a year-round ally, so their Pride post is more than corporate lip service.

submitted 15 days ago • 0 comments

As if I needed more reasons to love my favorite game:

submitted 20 days ago • 0 comments

I've been a Linux user for 15 years and it still brings me joy to shut down my computer by typing "shutdown 0"

submitted 20 days ago • 0 comments

It's amusing how often coding assistant AIs get basic language-dependent behavior wrong. It seems a lost cause to get them to reliably parse the behavior of pointer receivers in Go.

submitted 23 days ago • 0 comments

youtube.com/shorts/b8uyx...

submitted 25 days ago • 0 comments

EVERYBODY WANTS TO RULE THE WORLD EXCEPT ME is out today! www.amazon.com/dp/B0DMNHMKRJ/ This is book two of the DARK LORD DAVI duology, the thrilling conclusion!

submitted 26 days ago • 19 comments

"Thank you for your interest" might be one of the worst ways to begin an email communicating that a job application is proceeding.

submitted 26 days ago • 0 comments

Learning to set up Python virtual environments - so much AI-generated garbage. If you have found this post trying to find actually-helpful documentation, look here: documentation.ubuntu.com/ubuntu-for-d...

submitted 27 days ago • 1 comment

Benchmarked Golang's built-in map against a special-built statically-sized Hash Set implementation. It's surprising how much overhead (both in processing and memory) there is in general-purpose code, but it sure is nice to not have to solve the problems of DIY. willbeason.com/2025/05/21/p...

submitted 32 days ago • 0 comments

Clicking on links in form emails sent by admin people amuses me because of how often those links are broken. I always let them know, but I get the feeling I'm the only one who does this.

submitted 33 days ago • 0 comments

Well, it's nice to find out that an algorithm problem I was failing to solve all evening is apparently used as a term project.

submitted 33 days ago • 0 comments

Today's reading nook

submitted 39 days ago • 0 comments

Relaxing in Georgia, reading outside with a friend's goofy dog

submitted 40 days ago • 1 comment

Here’s my GenAI take, now and forever: if you have AI write a brief for you, you’re letting it make choices that you don’t even realize it’s making. Where do you start telling your story? Which names do you use for your characters? It’s a mistake to cede control over those decisions.

submitted 46 days ago • 2 comments

The Expanse TV show is a great adaptation, but you're missing over half the story and ideas of the series.

submitted 43 days ago • 0 comments

I really love this conversation, even if not every element worked perfectly for them. If something I write can spark a frank and robust discussion, that's as rewarding as five-star, no notes praise. Thank you, Lilly and Sara, for reading and talking about Down in the Sea of Angels!

submitted 45 days ago • 1 comment

Well neat, the naive solution (using maps) to calculating string permutations is about 100x slower than an array-based solution for long strings. willbeason.com/2025/05/08/o...

submitted 45 days ago • 1 comment

In an absolutely devastating ruling, the Supreme Court has allowed the transgender military ban to go into effect. This is shocking - the ban explicitly called transgender people dishonorable for being transgender. It is as clear as day soaked in animus. www.supremecourt.gov/docket/docke...

submitted 47 days ago • 865 comments

The soundtrack to Monster Train is a masterpiece (by Jordan Chin). This track is an incredible example of leitmotif. youtu.be/48-gflWLeBw?...

submitted 45 days ago • 1 comment

Researchers Emily Bender and Alex Hanna argue the industry is surrounded by a parallel bubble of hype.

submitted 47 days ago • 0 comments

If you don’t understand why GenAI hallucinates so often, and most people don’t, read this: garymarcus.substack.com/p/why-do-lar...

submitted 48 days ago • 8 comments

Learning to write performant Rust as part of interview prep. It feels good to regularly be writing code 5x-10x faster than generative AI. Not that GenAI hasn't been helpful; it's that it can really only help when I already know what I want to do. (So like, using Box and dyn properly)

submitted 47 days ago • 1 comment

The business model is literally "this site has a high trustworthiness in Google search results, so we bought in and we'll spam the domain with ad-laden click bait until the algorithm catches up"

submitted 50 days ago • 1 comment

Wrote a technical deep dive into a common programming interview question: Write a function that returns true if all characters in a passed string are unique. While the initial solution is obvious, it has neat properties when you try to optimize it. willbeason.com/2025/05/02/p...

submitted 50 days ago • 0 comments

Writing another short story. Got a few hundred words done this morning! Going to try and finish it this week. Aiming for 2,000-3,000 words.

submitted 55 days ago • 1 comment

LLMs are nothing more than models of the distribution of the word forms in their training data, with weights modified by post-training to produce somewhat different distributions. Unless your use case requires a model of a distribution of word forms in text, indeed, they suck and aren't useful.

submitted 59 days ago • 56 comments

Finished a little flash fiction story today (~600 words): willbeason.com/2025/04/24/a...

submitted 59 days ago • 0 comments

👉 👈 Meta announced that they're changing their models to reduce "left-leaning [political] bias"--that means leaning them to the political "right". Lots to unpack about what that might mean. So I ran a quick "shot in the dark" study...and found a political right bias in Meta models. Some notes.🧵

submitted 61 days ago • 28 comments

People spend a lot of time wondering why the left is worse at using independent media to control the national narrative. A useful exercise is to ask why the left is also worse at selling snake oil and pyramid schemes, because I think the reasons are basically exactly the same.

submitted 62 days ago • 1010 comments

Me: "I haven't run enough this week." Also me: has run over 15 miles this week Wild how the unattainable has become "not enough"

submitted 65 days ago • 2 comments