220 karma · joined December 10, 2017
- Next.js
- OpenAI's embeddings and GPT endpoints
- Postgres with pgvector (on neon.tech)
- Tailwind
- tRPC
- Vercel for web hosting
- Google Cloud products for data pipelines (GCS, Cloud Tasks)
Agreed it could be faster for uncached queries. The embeddings retrieval itself is actually pretty fast (uses pgvector). However, I found that having a LLM rerank results + generate summaries related to the search query made results more useful, which is what accounts for much of the latency.
Maybe I should make that a user-customizable setting!
This person was diagnosed in her 30ies. Her GP suggested that, along with maintaining a good diet, she learn proper swimming technique and start swimming daily. She has been swimming daily for the last 40 years and has essentially not suffered from any symptoms for decades, except when she's had to take breaks due to either travelling or other health reasons.
Obviously, no A/B testing was done here, and I do not know how this could translate to someone else. But maybe, just maybe, it could help someone reading this?
Also, swimming is a rather pleasant form of exercise in my experience (though I do not suffer from AS). So assuming there are no counter-indications, folks may want to give it a try!
While I assume that what is described in the article might have happened with other verticals, I would be curious to see how this will play out in the larger emailing ecosystem given the current hype & aggressiveness of the commercial practices of at least some of the players in the cryptocurrency field (as attested by others here).
"‘Nuff said"
(p. xliv)
Alex Kozinski truly is not a typical judge.