Matt Hodges
banner
matthodges.bsky.social
Matt Hodges
@matthodges.bsky.social
Trying to use computers to elect Democrats / Previously Joe Biden / Previously Hillary Clinton / Austin, TX / matthodges.com
I deleted Slack off my phone. Its place on my Home Screen is now occupied by a diaper and feeding logging app. I love that.

matthodges.com/posts/2025-1...
Dad
I love my daughter so much. I love my wife so much.
matthodges.com
November 2, 2025 at 6:52 PM
I wrote about why the marketing tech standard of hashing PII is silly, and I abused DuckDB into being a hashing mill.

matthodges.com/posts/2025-1...
The Privacy Theater of Hashed PII
It takes a 2020 Macbook Air four hours to hash every North American phone number
matthodges.com
October 19, 2025 at 5:14 PM
This week I achieved a few milestones: I hit 1,000 rides on Peloton, I weighed in at my goal, and I crossed five years of logging all this health data. I wrote a little bit about it here.

matthodges.com/posts/2025-1...
Tracking Five Years Of Health Data
A thousand rides, a stack of apps, and actual motivation to get in shape
matthodges.com
October 12, 2025 at 7:05 PM
I’ve had Aristotle in my Goodreads queue for a decade. ChatGPT finally got me reading philosophy.

matthodges.com/posts/2025-0...
ChatGPT Got Me Reading Plato
Notes, Prompts, and Noble Lies: A New Way to Jump Into Old Philosophy
matthodges.com
September 23, 2025 at 10:01 PM
Reposted by Matt Hodges
I wrote about what Gödel, Escher, Bach teaches us about the impossibility of solving AI prompt injections.

matthodges.com/posts/2025-0...
Music to Break Models By
Gödelian limits of prompt-safe AI
matthodges.com
August 27, 2025 at 1:16 AM
I wrote about what Gödel, Escher, Bach teaches us about the impossibility of solving AI prompt injections.

matthodges.com/posts/2025-0...
Music to Break Models By
Gödelian limits of prompt-safe AI
matthodges.com
August 27, 2025 at 1:16 AM
I wrote about using LLM embeddings to normalize user data and how that could apply to political campaigns who want to better understand their supporters. No chatbots involved!

matthodges.com/posts/2025-0...
Using LLM Embeddings to Normalize User Data
How cosine similarity in embedding space can power high-quality normalization.
matthodges.com
August 2, 2025 at 7:29 PM