Funny, I feel the same way about humans.
HN user
maccaw
How did he pull millions out of the economy? How is his money not in the economy?
Fwiw my experience has been the exact opposite. It is by far away, the best serverless database I’ve used.
Why should we make it more efficient? Why not make more energy from clean sources first?
they were favoring elon
No, and that's not what the article says either. They were just tracking how well his tweets were doing versus others. They were not favoring Elon.
Which parts do you reject?
I've been using SuperWhisper since the start and it's totally changed my workflow. I actually bought a mic because of it and I now mostly narrate my emails and other long form texts. The larger models are very accurate.
I know Neil is working on improving the speed of it using CoreML, which is I think the next big leap forward. Congratulations Neil!
The trouble with coupling the two is that now I have to either trust some third party db service, or some custom db.
The DB is so fundamental, and so hard to get right / scale, that it's hard to trust a startup to do it well.
We use Firebase and even with all the resources google has this has a lot of issues (especially with the client libs).
I have a genuine appreciation for how Linear has built this. We have had to build something similar for our note taking application (Reflect). It is very tricky to do and I wish there was more research on this.
In my opinion, what we need is:
1) A client-side performant SQLite database that supports live queries. I.e. you can automatically re-render the page when the queries change. That way your database can drive the UI and be the source of truth in regards to what's displayed on the screen.
2) A separate realtime syncing protocol that syncs database state to client state.
And ideally this is all open source, and that these two endeavors are not coupled tightly.
[1] Wa-sqlite is the best (imo) client-side db - better than than the official Sqlite WASM build (for now) because it had a indexeddb fallback for browsers that aren't the cutting edge Chrome.
[2] cr-sqlite is an interesting project using CRDTs to sync state around. However I still believe that for many production use-cases you want a ultimate server source of truth.
[3] Replicache is still the best closed source solution I know of.
[1] https://github.com/rhashimoto/wa-sqlite [2] - https://github.com/vlcn-io/cr-sqlite [3] - https://replicache.dev/
We've made it pretty abstract, so it should play nicely with langchain etc. We're all open-source and using open protocols like OpenAPI.
You make a good point though. We should write a langchain plugin to our API.
Learn TypeScript, Next.js, Vercel, and Tailwind. That's my advice.
Sign up for Vercel and play around with the sample apps.
This is great! See also Voy, A WASM vector similarity search written in Rust:
Have you two seen this?
We make both Whisper and GPT-4 available in https://reflect.app
We have all of the above. Export anytime. It's part of our company values.
Then you can export your notes to HTML or Markdown.
Yup, audit details are here: https://reflect.academy/security-and-encryption
I totally agree. In fact, that's the entire reason we didn't raise venture capital, because that would have forced an 'exit' (which typically means an acquihire/getting shut down). I'm sick of seeing otherwise good businesses ruined because they raised VC without a good reason to.
Our goals are modest, we're just trying to create a sustainable business. We're currently half-way towards profitability, and are on track to be profitable by the end of the year.
Regarding export formats, you can always export your notes into markdown or HTML so you shouldn't feel too locked in.
I love this! You should be able to do both of these today - we let you save custom prompt templates.
Woah, cool to see Reflect on HN.
I started Reflect a few years ago. We're building it as a lifestyle company. There's only four of us. We bootstrapped for a while, and then did a $1m crowdfund. Now our goal is to get to profitability, and we're half-way there.
Product wise, our goal is to make something beautiful, fast, secure - that just works out the box. We're end-to-encrypted, which I think this crowd will appreciate.
Happy to answer any questions.
Love it - also a huge fan of In our Time.
V2 would be:
- Transcribe through Whisper - Semantic search
Well technically everything is deterministic.
You don't think solving every disease is a significant upside?
Under persistant storage options:
The author's last-years underrated reasons to be thankful was also excellent: https://dynomight.net/thanks/
I was rejected with clearbit.com
Was told the business required a degree of enterprise sales that my skills weren't suited for. It's now worth $350m (on paper at least).
That said, no hard feelings. They see so many applications - they're bound to make the wrong decision time to time.
Do you also think you should be compensated by OpenAI for all the blog posts you've written that went into GPT3's training?
100%
Absolutely! You're more than welcome. Thank you.
I think because the answer here is using Next.js with hosted backend solutions like Firebase, Supabase, Pusher, etc.