rotexo · HackerTrans

há 3 mesesdiscuss

I love seeing this. Klondike with a physical deck was my “I need to do something with my hands or I’ll go crazy” release valve during the pandemic after work. I thought about trying to build something like this, but it was far outside of my technical ability. I settled for a simulator that would run toilet solitaire so I could see the distribution of cards remaining after games with a well-shuffled deck.

rotexo

há 3 mesesdiscuss

E4B is pretty good for extracting tables of items from receipt scans and inferring categories, wish this could be called from within a shortcut to just select a photo and add the extracted table to the clipboard

rotexo

há 6 mesesdiscuss

Paul Rudd’s computer (~2009?) was to me probably the most accurate prediction regarding genAI (https://youtu.be/a8K6QUPmv8Q)

rotexo

há 6 mesesdiscuss

Interesting that enforcing an ontology seemed to result in reduced performance. (https://youtu.be/hLDKW3-OKUA) I guess developing an ontology based on the extracted graph will be an important area for future work

rotexo

há 8 mesesdiscuss

I thought most asteroids were basically just gravel piles loosely held together by internal gravity?

rotexo

há 8 mesesdiscuss

Can you describe this more? Is “edge_case” a key in the structured output schema?

rotexo

há 10 mesesdiscuss

The stacks of tablets were because of DRM (a cybersecurity method to manage the threats of data exfiltration, “Digital Romulan Management”)

rotexo

há 6 anosdiscuss

That last paragraph hits home. I was always ‘good at school’ (even though I spent a lot of effort studying to maintain that status), up until I got to my PhD program and I couldn’t just ‘regurgitate simple models’ and succeed anymore. I got my PhD in the end, but I’m not as successful as many of my peers have been. Acceptance that I am not an elite scientist and committing to the work anyways was the thing I learned in grad school that has helped me in industry so far.

comments