Show HN: ChatGPT UI for rabbit holes • Hacker News Board

Re: Show HN: ChatGPT UI for rabbit holes

#101

Post by popalchemist » Thu, May 30, 2024, 7:42 PM UTC

igorguerrero wrote: post #99

This UI is fantastic! Was the name "delve" AI generated? I just cannot believe a human uses that word...

It's obviously a reference to the meme going around about how ChatGPT uses that word inordinately frequently.

Re: Show HN: ChatGPT UI for rabbit holes

#102

Post by horacemorace » Thu, May 30, 2024, 7:57 PM UTC

Super cool!

What’s the logic used to determine what words are links? I found that it was possible to feed it new vocabulary and that it would turn those into links

Re: Show HN: ChatGPT UI for rabbit holes

#103

Post by rexreed » Thu, May 30, 2024, 8:17 PM UTC

Which LLM is this using because it is providing hallucinated facts that differ from ChatGPT 4o on the latest model?

Re: Show HN: ChatGPT UI for rabbit holes

#104

Post by rexreed » Thu, May 30, 2024, 8:18 PM UTC

instagraham wrote: post #65

This could replace the hours I spend on Wikipedia. Hope it's not too expensive to run.

But not as accurate. A few quick queries and already it's providing misinformation. As much as Wikipedia has problems with truthiness in some instances, it's not nearly as bad as misinformation-ridden GPTs.

Re: Show HN: ChatGPT UI for rabbit holes

#105

Post by gsuuon » Thu, May 30, 2024, 8:31 PM UTC

This is cool, it's almost like a wiki you can talk to. I also wanted to make a thread-based UI for LLM chat since I realized that's how I typically interact with them (almost like git branches) but hadn't gotten around to it yet. Neat to see others are interested in branching conversations as well!

Re: Show HN: ChatGPT UI for rabbit holes

#106

Post by thelittleone » Thu, May 30, 2024, 8:33 PM UTC

wiremine wrote: post #78

Taking a step back: The UX/UI for LLMs in general are very immature. We're in the very early days of to best interact with these tools. We need more experimentation like this to help figure out what works, and doesn't work. Kudos!

Hopefully HCI improves enough that interfaces will be a last resort.

Re: Show HN: ChatGPT UI for rabbit holes

#107

Post by peperunas » Thu, May 30, 2024, 8:38 PM UTC

This is an outstanding idea, I love it!

It would be great if you could support Ollama (or an OpenAI compatible endpoint) for private LLMs.

Re: Show HN: ChatGPT UI for rabbit holes

#108

Post by andreygrehov » Thu, May 30, 2024, 8:41 PM UTC

I see there are many comments about Wikipedia. I can't find the link, but many years ago (like 7+), there was a concept Wikipedia redesign that proposed the same UX. This was a "marketing" project of a non-Wikipedia related designer. Does anybody remember?

Re: Show HN: ChatGPT UI for rabbit holes

#109

Post by jmkni » Thu, May 30, 2024, 8:43 PM UTC

Matticus_Rex wrote: post #62

jmkni wrote: post #60

Earlier quoted context omitted.

what's the joke?

ChatGPT overuses the word "delve" in its writing. Search your email for "delve," and look at how common it starts becoming (esp. in marketing emails) around the time ChatGPT takes off.

Ah gotcha lol

Re: Show HN: ChatGPT UI for rabbit holes

#110

Post by ajdegol » Thu, May 30, 2024, 8:51 PM UTC

Branching conversations are great for a whole bunch of reasons. I posted a demo of a prototype: https://x.com/ajdegol/status/1788689011302682657

And Jake Collins just announced he’s open sourcing an obsidian plugin which has a ton of features: https://x.com/JacobColling/status/1795462258258002255