Don Richards
@donrichards.bsky.social
Sr. Software Engineer at Johns Hopkins University. Turning code, data, and library science into open knowledge ecosystems because the best ideas deserve a platform, not a paywall. Moonlighting as a pretend data scientist.
It just occurred to me that there is roughly ≈ 4,600+ metric tons of SpaceX satellites in the sky. These are expected to have a lifespan of roughly 5 years. A quick search and all I'm seeing researchers modeled the release of 10,000 tons of aluminium oxide annually by 2040. Is this enough? #starlink
September 12, 2025 at 2:52 PM
It just occurred to me that there is roughly ≈ 4,600+ metric tons of SpaceX satellites in the sky. These are expected to have a lifespan of roughly 5 years. A quick search and all I'm seeing researchers modeled the release of 10,000 tons of aluminium oxide annually by 2040. Is this enough? #starlink
When you don't trust live demos for a presentation at a conference so you record the demo and add a comment to your slides
> This Live demo was brought to you by an MP4 file on my desktop. Enjoy!
> This Live demo was brought to you by an MP4 file on my desktop. Enjoy!
July 9, 2025 at 6:54 PM
When you don't trust live demos for a presentation at a conference so you record the demo and add a comment to your slides
> This Live demo was brought to you by an MP4 file on my desktop. Enjoy!
> This Live demo was brought to you by an MP4 file on my desktop. Enjoy!
Traefik’s automatic wildcard certs once made setting up dev environments a breeze but now it seems they’ve lost the certificate integration battle. What does that mean for users? “Sorry, they won” isn’t the commit message anyone wants to write. #traefik #certs #ssl #letsencryptfail
March 25, 2025 at 3:16 PM
Traefik’s automatic wildcard certs once made setting up dev environments a breeze but now it seems they’ve lost the certificate integration battle. What does that mean for users? “Sorry, they won” isn’t the commit message anyone wants to write. #traefik #certs #ssl #letsencryptfail
February 3, 2025 at 6:24 PM
January 30, 2025 at 6:02 PM
Project Silica uses quartz glass for indefinite storage. Based on this article I speculate it could enhance AI by ensuring data integrity and helps build more sustainable infrastructure for AI development. Towards Sustainable Cloud Archival Storage in Glass - doi.org/10.1145/3708...
January 28, 2025 at 9:51 PM
Project Silica uses quartz glass for indefinite storage. Based on this article I speculate it could enhance AI by ensuring data integrity and helps build more sustainable infrastructure for AI development. Towards Sustainable Cloud Archival Storage in Glass - doi.org/10.1145/3708...
On my very first prompt today, Claude hit the max limit... on a paid account 😅.
Alright, back to ChatGPT I go if you can’t handle the heat, I know who can. 🔥
Anyone else bouncing between AI tools like it’s musical chairs? 😂 #claude
Alright, back to ChatGPT I go if you can’t handle the heat, I know who can. 🔥
Anyone else bouncing between AI tools like it’s musical chairs? 😂 #claude
January 7, 2025 at 3:06 PM
On my very first prompt today, Claude hit the max limit... on a paid account 😅.
Alright, back to ChatGPT I go if you can’t handle the heat, I know who can. 🔥
Anyone else bouncing between AI tools like it’s musical chairs? 😂 #claude
Alright, back to ChatGPT I go if you can’t handle the heat, I know who can. 🔥
Anyone else bouncing between AI tools like it’s musical chairs? 😂 #claude
An intelligent person may express something simply when they deeply understand a topic. Another expert would see this as a sign of mastery, while the same comment from a child wouldn’t be seen as advanced knowledge. The issue is that AI doesn’t take this “common sense” approach in interactions.
AI still lacks “common” sense, 70 years later
What’s obvious to people still isn’t always obvious to machines
open.substack.com
January 6, 2025 at 8:16 PM
An intelligent person may express something simply when they deeply understand a topic. Another expert would see this as a sign of mastery, while the same comment from a child wouldn’t be seen as advanced knowledge. The issue is that AI doesn’t take this “common sense” approach in interactions.
I'm liking the new layout for ChatGPT.
December 11, 2024 at 8:19 PM
I'm liking the new layout for ChatGPT.
