matt_holden

Interesting point. OpenAI has said they don't train on files uploaded via the API (like the Assistant API), but unclear what the policy is for documents in GPTs.

Either way, the signal they could get from understanding what KINDS of documents builders/users want to do better retrieval on is probably quite valuable.

I also wonder how user file uploads will interact with copyright law and the new Copyright Shield from OpenAI.

E.g. if a user uploads the full text of Harry Potter to a GPT, you could argue the model output is fair use but unclear how courts will interpret that.

LLMs are already a sort of "copyright blender" that aggregate copyrighted inputs to produce (probably?) "fair use" outputs. With the foundation models, OpenAI can decide what inputs to include in training. But with custom GPTs, users can now create their own personal copyright blenders just by uploading a PDF :)

Cool, makes sense!

It seems like you're heading in the right direction on packaging/pricing.

Charging for services you manage, more team members, and security features bigger cos care about all make sense to me.

I previously looked at your Pro extensions and some seemed nice to have (UniqueId, emoji), but it wasn't enough to get me to pay, and I rolled my own versions of those instead.

Building multiplayer sync is really tricky, so providing that as a managed service seems valuable and worth paying for. I have no problem paying for services like Clerk.com or Liveblocks.io that involve managed "servers" I don't want to build/run vs a purely client-side library.

I recently discovered Liveblocks, and think their positioning/marketing/docs are great -- the fastest way to turn any SaaS app multiplayer.

I wonder if you could do something similar but focused on all the complexities of text editing (esp since it's such a core primitive for LLMs now). Just a thought - excited to see where you take it!

Congrats on the launch, and great you'll be dedicating more focus to Tiptap!

If anyone wants to see another example of an app using Tiptap, I built https://duckbook.ai as a fun/free solo project (it's a Notion-like SQL notebook built on Tiptap, DuckDB, and GPT-4).

I evaluated a bunch of editors (ProseMirror, TipTap, Remirror, Slate, Quill, Draft.js, Lexical) and went with TipTap.

I've been generally happy with it, the APIs/docs are good. ProseMirror is super powerful, but large/slow to work with. Tiptap helped me launch a product much faster, but still pop the hood to the ProseMirror layer in a few places where needed.

Main request is to keep the tiptap/core project free and MIT licensed, even if you raise prices on other pieces (like the managed collaboration server).

Comments like this remind me of the old-timers from IBM saying "but wait, we invented the PC! and the cloud! and..."

Gotta put products in the market, or it didn't happen...

Look at Snapchat: https://techcrunch.com/2023/02/16/snapchat-announces-750-mil...

Just above 700m MAU. So yeah, probably aimed at their direct competitors in social.

Sam doesn't have much financial upside from OpenAI (reportedly, he doesn't have any equity).

And he wrote about the risk in 2015 months before OpenAI was founded: https://blog.samaltman.com/machine-intelligence-part-1 https://blog.samaltman.com/machine-intelligence-part-2

Fine if you disagree with his arguments, but why assume you know what his motivation is?

Also: Figma, Linear, Notion, Slack, Superhuman, Pitch, Loom. All Electron.

I’ve never understood the hate Electron gets on HN, when it powers so many great products that users love.

When I was a product leader at Dropbox back in 2015, we chose to build our Gmail integration on the InboxSDK and we're generally happy we did.

It let us focus on our core user experience, without worrying about all the DOM hacking, Gmail variants, constant version updates, etc.

If you want to build cool experiences in Gmail, highly recommended.

Is this true though?

I can send $1000 USD to someone in France who wants EUR for $7 with Western Union.

How does Bitcoin or Ethereum let me do this for less?

The focus on compactness (or other geometric properties) as a goal is misguided.

The districts don't exist to look pretty or have nice geometric properties. They exist to translate the will of the voters into a representative government.

538 compared maps gerrymandered for various goals [1]. Prioritizing compactness has the effect of a) skewing political power toward rural areas away from cities, and b) reducing the expected number of non-white House members (for similar reasons).

The problem with compactness is it doesn't take into account the political geography of where people live. Increasingly, Democratic majorities are concentrated in cities, while Republican majorities are outside of them. Drawing compact districts has the effect of packing Democrats in fewer (urban) districts, resulting in more wasted votes. This gave the Republican party about a 30 seat advantage in the 538 map, compared to a more proportional one.

The Senate and Electoral College already unfairly skew political power toward rural voters. Compact maps would do the same in the House.

The efficiency gap (your #3) is a much better measure of partisan fairness to focus on than compactness.

[1] https://projects.fivethirtyeight.com/redistricting-maps/#alg...

The 2015 federal transportation bill allocates $48B to transit over 5 years and $205B to highways.

I've never understood why people frequently describe transit spending as a "government subsidy" but would never say the same thing about the much larger highway spending.

If you don't insist on adding tolls everywhere to make highways break even, why hold transit to a higher standard?

The unpopularity of annuities seems like a direct consequence of loss aversion, which says losses are more psychologically powerful than equivalent gains. [1]

If I trade $100k in savings for an annuity, I'm starting off at a deep "loss", which becomes a "gain" only at some point in the uncertain future. Not surprising that most people won't take that bet.

[1] https://en.wikipedia.org/wiki/Loss_aversion

(Hi all, I'm the PM for the Dropbox desktop client team.)

I just wanted to let you all know that we take any claims like this really seriously. There aren't any known bugs on the Dropbox side that would cause this, and unfortunately there are potential causes such as hardware errors, filesystem corruption, and other OS issues (including those like http://www.phoronix.com/scan.php?page=news_item&px=MTIxN... which another poster pointed out) that can corrupt data or create zero byte files.

Nevertheless we will continue to look into this just to be sure, and we also work hard to find ways for Dropbox to shield users even when the OS, disk or other components fail (our undelete/file revisions and Packrat are among these).

No, you'll only get prompted on desktop/mobile the first time you link your account. If you already have a linked desktop or mobile app, it will continue to work without a prompt.

You can also see all of your currently linked devices here: https://www.dropbox.com/account/security