nerdtangent.bsky.social
- Nerdy things
- Writing things
- Random things
- Published Author
45 posts
5 followers
19 following
Regular Contributor
Conversation Starter
comment in response to
post
This is why I refer to commands in the CLI as incantations. It is straight up arcana at this point.
comment in response to
post
So Google Images is currently still just an image search which means you can still use negative key words to filter out AI images. No clue how much longer that'll last, but for now, adding things like "real", or "not generated by AI" should provide a sufficient filter.
comment in response to
post
Nope that's it. Just software. Only real difference is this software is geared towards generalized tasks within a given domain.
comment in response to
post
If you solve for patterns, then those patterns easily translate and extend to whatever tool chain has you hot and bothered at the time. Further, if your pattern solves for the 80%+ you don't need to maintain nearly as many patterns.
comment in response to
post
I imagine they will once they evaluate their way forward. Give it time.
comment in response to
post
comment in response to
post
Or that DevOps and SRE are things you can buy and not cultural shifts that require significant effort to achieve.
comment in response to
post
The modern data stack isn't modern and that's part of the problem. The sources tend to be 10+ years of crap. If you don't have this problem, you don't need all the shiny tools. You need Postgres. Slightly boring, but does what it says on the box.
comment in response to
post
Been thinking about this for years. I like your approach, minus the Rust for backend I think. I don't have an opinion there because I still need to sit down and get my hands dirty with it.
comment in response to
post
Don't forget the duct tape!
comment in response to
post
comment in response to
post
Agree 100%. Biases are super tricky since all data has biases built into it by design. I still believe we can do better when it comes to building representative datasets. The annoying thing is that synthetic data is more of a thing now. Which is designed off of biased data.
comment in response to
post
Woo! Glad you were able to make some time. You've got this.
comment in response to
post
This is the way.
comment in response to
post
This was quite some time ago, but still very cool :)
comment in response to
post
Honestly, you sometimes have to make time. Block off 2-4 hours per week. Or even 2-4 hours per month if you're super busy. It's a side project, so slow and steady wins the race.