@DhravyaShah Well, you know I’ve been asking for selecting multiple specific drives in OneDrive. Right now, OneDrive pulls the entire OneDrive.
At the same time, the console limits to 500 items.
Most people will have far more than 500 items in their drives.
@_prasanna_ap@DhravyaShah Thank you, Prasanna. I've checked the dm and saw the email you sent. I just responded.
I appreciate you guys looking into this.
Any update on when selecting specific multiple folders will be coming to the OneDrive connector? I've already discussed this with Dhrayva.
I just saw that Microsoft has a OneNote plugin for ChatGPT in beta and it seems that it would read but also create pages, sections, etc.
Which is hella dope and would have made me consider returning to OneNote,
except that agents can’t read it as well as Obsidian
Sigh. Oh well
@DhravyaShah Thanks. Yeah at first the issue was trying to figure out how to select multiple specific folders to sync. Then the issue became selecting a folder would not initialize a sync and the UI keeps a popup u saying to sign into google even though I’m signed in. Appreciate it.
@ibocodes I’m also kinda like “this all sounds very cool. But what does it all mean and how do I use it?”
And so far every “Jev simplified explainer” vid is just rambling about how cool it is without actually explaining HOW it works or HOW to actually implement it.
I still don’t fully understand Jev, but I’m thinking if it is super duper at decisions, then one could build something like… a model router?
Something that can immediately switch between different AI models based on complexity or something.
Is that a viable use case?
I keep trying to watch people explain what typesafe and Jev is and either my brain is not too good or people claiming they can simplify it are bad at simplifying it.
Or maybe a mix of both.
@thisiskp_ Is Jev’s “yes/no” based on how you specifically train it or does it have its own built in understanding?
During sign up, it gives an example of is a hotdog a sandwich?
But how does Jev know the answer? How can it answer with confidence?
Today we're rolling out Projects in Claude Code on desktop and web.
A project is one conversation with Claude. It splits the work into threads itself, runs them as parallel cloud sessions, passes context between them, and keeps going when you leave.
In beta for select users.
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.