I swear this is the year I go to #rustconf @rustconf.com
04.06.2025 11:03 β π 2 π 0 π¬ 0 π 0@jodavaho.io.bsky.social
PhD, Computer Science. Robotics and Planning systems for earth and elsewhere. https://josh.vanderhook.info https://jodavaho.io https://twitter.com/jodavaho https://mastodon.social/@jodavaho I write code for pilots right now at https://flightscience.ai
I swear this is the year I go to #rustconf @rustconf.com
04.06.2025 11:03 β π 2 π 0 π¬ 0 π 0I like this. Many don't.
Use structured logging - you can filter /query so you're not scrolling for ages and don't upset the other devs who want to find *their* code except *your code* is spamming all over it.
Great article! Glad to see SDL + Linux are goto options for gamedev.
20.05.2025 19:08 β π 5 π 0 π¬ 0 π 0I'm nearing the end of a week long "quick refactor" and I feel like a dumbass. Oh well.
15.05.2025 23:58 β π 0 π 0 π¬ 0 π 0Any sufficiently reductive analysis is indistinguishable from a takedown.
Someday we'll see that we ourselves think much less than we think, LLMs are probably a lot closer to us than we care to admit, which is why they are so dumb and overconfident.
The other side would say education is fine indoctrination is not. There is a distinct lack of diverse politics in higher education.
But honestly the most divisive politicization is not from tech classes, that should be obvious.
Come on, people have been wordsmithing their grants, publications, results, and proposals since well before the days of Galileo and Copernicus.
Stop pretending to be injured and just coddle the ruling class as we always have, then go good work. The bigger stink, the more censoring.
Unfortunately "keep politics out of science" cuts two ways. IMHO it's best to comply and as quickly as possible continue funding the grants that would be acceptable to the admin, vs trying to take a stand and burning down the whole thing.
02.05.2025 11:37 β π 0 π 0 π¬ 1 π 0Your regular reminder that NSF grants built the internet and helped the U.S. dominate tech.
02.05.2025 11:02 β π 1215 π 480 π¬ 35 π 7If given the choice between more work done and fewer employees, most healthy businesses choose productivity. But it might slow hiring.
02.05.2025 10:55 β π 0 π 0 π¬ 0 π 0watercolour of a battle between british and french sailing ships.
1804. Oceans are now Battlefields.
In the South China Sea a British convoy worth almost Β£1bn today is spotted by the French.
It should be a massacre.
Instead, Commodore Nathanial Dance is about to defeat a French battle squadron using some paint and the most overplayed hand in #navalHistory. /1 π§΅
I'd watch this 6x a week sometimes. It just felt like home for some reason.
26.04.2025 02:28 β π 0 π 0 π¬ 0 π 0I've found a person with proper ownership and potential upside will probably work more than a "normal" job, with N=1 experience from myself. It's *hard* to work long hours if you're not engaged and telling yourself the right story about the work.
25.04.2025 11:47 β π 2 π 0 π¬ 0 π 0If Wikipedia had released the first chat AI we'd be heralding a new age of knowledge, democratic access & achievement.
But listless programmers built it first then decided to automate their own jobs. So here we are, white collars, worrying about the dangers of it all not one bit aware of the irony.
Got the itch to build something. Might go back to tcmaker.com, but last time wasn't a great experience.
The problem with these clubs is the unspoken rules. I want to go in and use a table to cut something but it's not the way everyone else does it.
What's potatoia.club?
01.02.2025 01:18 β π 2 π 0 π¬ 0 π 0I think I figured out why one core on this machine runs hotter than the others.
27.01.2025 14:22 β π 4 π 0 π¬ 0 π 0With some thrashing, I've started to chip away at the big 3 learning goals for 2025-2028:
* Quantum Computing
* RL for planning (with obligatory side quest into LLMs)
* CUDA and programming massive parallel processors
On QC: The @mitopencourseware.bsky.social courses are great!
A new LLM is like a new source of steel: Fascinating scientific discovery. But nothing that really affects anyone not using raw materials.
arxiv.org/abs/2303.10130 predicts 80% of jobs will have 10%+ LLM integration, 20% will have 50% integration. We're nowhere close. Go build!
We're researching vacation destinations for next year, and with the kids and school and all that, can really only swing one trip. I'd like to get satellite photos and data of the locations to get a real picture of the layout, amenities, location, congestion, etc.
Tips? What data would you use?
@mcfunley.com Really amazing presentation here: egoless.engineering
Thank you! Nailed so many things I felt but couldn't express at past jobs.
Eh, it's fun. Try it.
01.12.2024 22:19 β π 2 π 0 π¬ 0 π 0I just completed "Historian Hysteria" - Day 1 - Advent of Code 2024 #AdventOfCode adventofcode.com/2024/day/1
01.12.2024 14:12 β π 1 π 0 π¬ 0 π 0I'd like to learn basic chemistry. Catch is I only have time for audio books about 3 hrs a week while jogging or working out.
There are zero good beginner / college level chemistry books as audio books. Tips?
Nobody will ever give his admin credit for NOT having an unemployment disaster during inflation.
29.11.2024 23:22 β π 3 π 1 π¬ 0 π 0Was just chilling with family. I take at least 3 hours and do nothing but family stuff every day. It's great.
But I'd love to have 3 days straight to work every now and again. The Grad School sprint dies hard.
@wmk.io π
24.11.2024 18:58 β π 1 π 0 π¬ 0 π 0I am not an expert.
IMHO Politicization of science is good and natural w/ diverse viewpoints leading to diverse policy recommendations.
If you have single sided or echo chamber politicization, you get one sided policies and ostracize too many to be viewed as trustworthy. Breeds mistrust.
So have you built this rig or something similar?
23.11.2024 23:51 β π 0 π 0 π¬ 1 π 0