i started a newsletter momswhobike.substack.com
@hamel.bsky.social my guess is you have a POV here...
i want your strongly held opinions on pgvector vs specialized databased for vector embeddings.
at what point do teams switch to specialized alternatives? when is it worth it?
The final US Presidential vote count is out (yes, I know, rest of the world!). According to Cook Political Report analysis, the election was decided by just 229,766 votes across MI, PA and WI out of about 155.2 million cast nationally.
Alright #dataBS: for anyone using Databricks -
What do you mostly use it for? What made you choose the tool? Where do you find it solving your problems most?
I was a huge fan and early adopter of Snapchat Spectacles (RIP). So I like this directionally but am unimpressed so far. Also, the glasses feel very cheap. I'd pay more to not feel like I'm wearing glasses that feel like they came out of a 5 year old's birthday piñata.
- I have a purebred Aussie, Meta AI suggests she is a Border Collie. For anyone that knows my dog, this is hilarious.
- I connected the Meta View app to my Spotify account but I keep being told "I can't do that on your glasses" when I ask it to do anything Spotify related.
- "Spotify Tap", when mapped to Spotify, plays a random recommended song from Spotify. That's pretty useless imo.
Initial experience of Ray Ban Meta glasses:
- I couldn't install firmware update via a closed wifi network on my Pixel. The Meta view app itself was v unhelpful, Meta documentation not much better. Gave up, it worked fine on iPhone.
Not everyone realises how easy it is to run a local LLM these days. If you’re on MacOS with Homebrew:
> brew install --cask ollama
> ollama run llama3.2
That’s it.
Still capturing the incredible journey of my 22-year-old son Elliott, on his epic charity cycling adventure from France to Australia.
15,000 KM achieved 🚴♀️
The #dataviz part is the easy part
public.tableau.com/app/profile/...
I want to hear more about the desire for an alternative to SQL!
Cycling is 10 times more important than electric cars for reaching net-zero cities and 100 times more fun.
now we need to hear for which tool-- or in which cases it is value add!
I think this is a fair take. When I ran the data team at Replit we actually replaced a number of Airflow pipelines w Hex. I always felt like it wasn't the proper way to do things but Hex made it really easy.
why dismay? i am intrigued to hear more!
great take!
Most accurate description of so much data work is below: "most of us are counting imaginary fictions and somehow getting paid for it"