vimota

9 ms·

31.
by vimota 3y ago
Similar to you, for polling you only see new data not the deletion events so I can't delete embeddings unless I keep track of state and do a diff. To properly support that you/I would need effectively CDC, which gets more complex
32.
by vimota 3y ago
Thanks for writing this up! I'm working on a very similar service ( https://embeddingsync.com/ ) and I implemented almost the same as you've described here, but using a poll-based stateful workflow model instead of
33.
Using DuckDB to serve low latency OLAP queries(twitter.com)
1 points by vimota 3y ago|0 comments
34.
by vimota 3y ago
Two years ago I became really interested in optimizing testosterone levels naturally and one of the things that kept coming up in research was the impact of EDCs (Endocrine Disrupting Chemicals) on fertility, cancer, hormone health and test
35.
by vimota 3y ago
Although I'm optimistic about this, a line popped out in the blog post stuck out to me: "deliberately training misaligned models". I'm guessing they mean misaligned small or weak models, and misaligned in a non-dange
36.
by vimota 3y ago
There's a good chance this is in response to scrapping caused by them suddenly deprepcating the /follows + /followers API : https://twitter.com/thomasschulzz/status/1674881114773815296...
37.
Root Venture's Investment Memo for Zed(github.com)
1 points by vimota 4y ago|0 comments
38.
by vimota 4y ago
Unison lang has a lot of overlap with model: https://www.unison-lang.org/learn/the-big-idea/
39.
by vimota 4y ago
This is great, thanks!
40.
by vimota 4y ago
Appreciate the kind words, but I'm sure you could pick it up pretty quickly too :) The OpenAI docs are pretty helpful as a starting point: https://platform.openai.com/docs/guides/embeddings/use-cases
41.
by vimota 4y ago
I played around with that and pgtrgm ( https://www.postgresql.org/docs/current/pgtrgm.html ) a bit but unfortunately didn't have great results. I did do a bunch of manual data cleaning though, and also had some
42.
by vimota 4y ago
Yep that's true! I'd probably do something like that if I were starting again, but the ease of calling a few APIs is pretty nice. I feel like that alone will drive a lot of adoption of some of these platforms even if it can just b
43.
by vimota 4y ago
I didn't use the GPT3 autocomplete API in the end (though I did play around with it) but did use the embeddings API (which I believe is still considered part of the "GPT3" model, but I could be wrong!). I totally could! I thi
44.
by vimota 4y ago
To be clear, I didn't use the autocomplete GPT-3 API just the embeddings one! Pinecone has some good docs on it: https://docs.pinecone.io/docs/openai . Happy to answer any questions :)
45.
Using GPT3, Supabase and Pinecone to automate a personalized marketing campaign(vimota.me)
252 points by vimota 4y ago|62 comments
46.
by vimota 4y ago
Isn't satiety / ghrelin the mechanism (via GLP-1) for weightloss in semaglutide?
47.
by vimota 4y ago
>Similarly, if AI applications become democratized and accessible to all enterprises, not just the hyperscalers, then it is Nvidia who will be positioned to pick up the entirety of the long tail. How does this square with Cloud providers
48.
An SPA Alternative(htmx.org)
278 points by vimota 4y ago|138 comments
49.
by vimota 4y ago
Dictation (speech-to-text) is also pretty crappy compared to Google's.
50.
by vimota 5y ago
https://vimota.me/ :)
51.
by vimota 5y ago
Uh what, can you say more about the diabetes part ?
52.
by vimota 5y ago
Oh I get that, and agree! But despite that it still feels like a honeypot, centralizing every user's most important security info in one cloud service (read: honeypot). At least with Dropbox/iCloud sync you're relying on the
53.
by vimota 5y ago
I've been a huge fan of 1Password for almost ten years now, recommending it to friends and family, but like some of the comments mentioned it feels like the product is trying to move upmarket while dropping support for core features. I
54.
by vimota 5y ago
That's part of the point though.
55.
In-Depth Tactics to Seriously Boost Your Energy(lynalden.com)
2 points by vimota 5y ago|0 comments
56.
Why do we crave sweets, and how can we stop?(levelshealth.com)
2 points by vimota 5y ago|0 comments
57.
Medical researcher taps existing drugs for new uses(mishpacha.com)
1 points by vimota 6y ago|0 comments
58.
by vimota 6y ago
My diy lamp shade has been wrapping parchment paper around the LED corn bulb (making sure to leave space for airflow so it doesn't overheat!). It's not perfect but it helps diffuse the light and make it less intense to look at.
59.
by vimota 6y ago
If you're too lazy to build something like this out but like the idea I highly recommend bright LED lighting for the day (explained here: https://www.benkuhn.net/lux/ ). I bought two ~7000 lumens LED corn bulbs and
60.
How Fortune 500 Companies Avoid Paying Income Tax(investopedia.com)
3 points by vimota 6y ago|0 comments

More ›