John Edwards
@johnbedwards.io
seattleite, professional baseballer, amateur jogger, semi-pro bar trivia ringer. assistant director of data science @ the seattle mariners, opinions my own
i feel like such a sicko for getting this reference
November 7, 2025 at 8:35 PM
i feel like such a sicko for getting this reference
(i am neither of these things either, hence why i'm operating with weaker priors _and_ trying to listen to those who are, like actual professors)
November 3, 2025 at 11:18 PM
(i am neither of these things either, hence why i'm operating with weaker priors _and_ trying to listen to those who are, like actual professors)
imho, the best recommendations for learning approaches come from people are experts in both their subject matter _and_ the learning process itself, which neither 1) nor 2) are
November 3, 2025 at 11:18 PM
imho, the best recommendations for learning approaches come from people are experts in both their subject matter _and_ the learning process itself, which neither 1) nor 2) are
i'm quite skeptical of LLM based learning (open to being wrong about this! but a lot of the LLM code I encounter is really flawed and doesn't perform well. and i review a lot of code).
November 3, 2025 at 11:18 PM
i'm quite skeptical of LLM based learning (open to being wrong about this! but a lot of the LLM code I encounter is really flawed and doesn't perform well. and i review a lot of code).
2) people who learned how to code with just an LLM and have a functional, but quite flawed understanding of the language and its best practices, and have not yet discovered the limits of their flawed education
November 3, 2025 at 11:18 PM
2) people who learned how to code with just an LLM and have a functional, but quite flawed understanding of the language and its best practices, and have not yet discovered the limits of their flawed education
1) people who learned before LLMs and are capable of recognizing when an LLM makes a mistake (but may not register that for someone less experienced, they wouldn't catch that mistake if the LLM made it) or
November 3, 2025 at 11:18 PM
1) people who learned before LLMs and are capable of recognizing when an LLM makes a mistake (but may not register that for someone less experienced, they wouldn't catch that mistake if the LLM made it) or
and this is coming from the data.table guy,,,,
November 1, 2025 at 10:06 PM
and this is coming from the data.table guy,,,,