Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I used GPT to build a search tool for my second brain note-taking system

reasonabledeviations.com

Post reply on HN
Print view View original

91–100 of 103 posts

Re: I used GPT to build a search tool for my second brain note-taking system

#91

Post by dewey »

rolenthedeep wrote: post #90
dewey wrote: post #75

Earlier quoted context omitted.

I feel like this was the promise of voice assistants like Siri many years ago, turns out they currently still have issues turning on / off lights reliably.

The issue with modern voice "assistants" is that it's impossible to monetize them and still keep them useful. Something like this will probably never get built by a capitalist corporation because there's no way to turn a profit. That's why siri and Google are driving their assistants into the ground with "by the way"

I don't think this makes sense, it's a feature that makes their platform more valuable even if not directly monetized. There's many features like that.

If someone would nail voice assistants that would be a bit selling point for their platform and it's a good way to get people into their paid services too.

Not really sure what you are referring to with driving the assistants in the ground, from what I can tell it's just not very good tech ("one moment", "working on it", "there was an problem answering this query",...) and not related to any monetization strategies.

Re: I used GPT to build a search tool for my second brain note-taking system

#92

Post by PeterisP »

sowbug wrote: post #23
bbor wrote: post #20

Earlier quoted context omitted.

Wait is this a serious concern or a joke? I think asking a chat bot whether it’s admissible is the same as asking someone’s handwritten diary if it’s admissible.

Fortunately, US courts take these sorts of questions quite seriously. See, e.g., State v. Smith (Ohio 2009): "Even the more basic models of modern cell phones are capable of storing a wealth of digitized information wholly unlike any physical object found within a closed container." https://en.wikipedia.org/wiki/Carpenter_v._United_States is also a good read, as is https://en.wikipedia.org/wiki/Riley_v._California :…

Those are all 4th amendment cases for warrantless searches violating privacy.

The fifth amendment has very different criteria - with a proper warrant, your most private things are admissible evidence. You can't be compelled to testify but all your most private notes in a safe can be "interrogated"; in many situations you can't be compelled to testify against your spouse but any writings or recordings of what you said about him/her are valid evidence. There is no debate that all the contents of your computer or phone can be used, they definitely can, the cases you quote are disputed only because they are "worthy of the protection for which the Founders fought" which is the requirement for a warrant.

I'd say that fifth amendment is not a privacy law (which is the 4th, stating that you have all the privacy without a warrant and no privacy with one), the 5th is essentially an "anti-torture" law to prevent coerced confessions, and there is no reason why it would apply to some physical evidence like the data for a trained model on your device.

Re: I used GPT to build a search tool for my second brain note-taking system

#93

Post by kanyethegreat »

ilaksh wrote: post #72
kanyethegreat wrote: post #40

Earlier quoted context omitted.

Would you be able to do something similar with non-text data (eg. tabular data)? For example, could you give it a bunch of excel files and ask it do give you the total units sold for an e-commerce site?

What I think would make sense would be to do it in two stages, and embeddings probably aren't really what you want. You would want to parse the Excel files into a certain data structure, and quite possibly text completions could help with that. Put that in a database or some files that you could use Python data science tools with or something. Then use text completions to translate the natural language query into som…

thanks for the response. unfortunately, the discord is full :sad:

Re: I used GPT to build a search tool for my second brain note-taking system

#94

Post by Terretta »

qwerty456127 wrote: post #70

I don't imply any judgment (like good/bad) but I tend to suspect the major (not necessarily intentional) reason of all these "second brains" (I use too) to exist in the grand scheme of things is to be a high-quality input for AIs to learn.

Obsidian is offline static markdown files

If offline static markdown exist to be input for AIs to learn, then OK.

But for me, long before Obsidian, they're just notes, because the act of writing them causes you to process them "outbound" which is the same process you need to recall them later.

Re: I used GPT to build a search tool for my second brain note-taking system

#95

Post by vineyardmike »

rolenthedeep wrote: post #89
zirgs wrote: post #69

Earlier quoted context omitted.

And reports you to police if you happen to do something illegal?

What do you think you're adding to this conversation?

It’s a genuine concern. If you are constantly recording your life, and sharing it with a third party, you’re building an incredible paper trail of every minor infraction.

There’s a whole bunch of “crimes” that society just kinda ignored at scale. Eg underage drinking in college. People knowingly and willingly speed when it’s against the law. Imagine if your smart car automatically recorded and stored its speed and gps coordinates at all times? It’d be so easy for the government to automatically subscribe to that data and start sending automated tickets… nevermind all the worse things that can happen.

This data can be manipulated and abused by stalkers and hackers, abusive partners controlling their wives or kids, churches trying to guilt you into behaving differently, etc.

Re: I used GPT to build a search tool for my second brain note-taking system

#96

Post by lannisterstark »

rolenthedeep wrote: post #90
dewey wrote: post #75

Earlier quoted context omitted.

I feel like this was the promise of voice assistants like Siri many years ago, turns out they currently still have issues turning on / off lights reliably.

The issue with modern voice "assistants" is that it's impossible to monetize them and still keep them useful. Something like this will probably never get built by a capitalist corporation because there's no way to turn a profit. That's why siri and Google are driving their assistants into the ground with "by the way"

You'd think that they can probably support this with the ad-revenue they're getting...

Re: I used GPT to build a search tool for my second brain note-taking system

#97

Post by rolenthedeep »

vineyardmike wrote: post #95
rolenthedeep wrote: post #89

Earlier quoted context omitted.

What do you think you're adding to this conversation?

It’s a genuine concern. If you are constantly recording your life, and sharing it with a third party, you’re building an incredible paper trail of every minor infraction. There’s a whole bunch of “crimes” that society just kinda ignored at scale. Eg underage drinking in college. People knowingly and willingly speed when it’s against the law. Imagine if your smart car automatically recorded and stored its speed and gp…

> If you are constantly recording your life, and sharing it with a third party

Which is why the very first sentence of my post includes the phrase "self-hosted"

Re: I used GPT to build a search tool for my second brain note-taking system

#98

Post by sowbug »

PeterisP wrote: post #92
sowbug wrote: post #23

Earlier quoted context omitted.

Fortunately, US courts take these sorts of questions quite seriously. See, e.g., State v. Smith (Ohio 2009): "Even the more basic models of modern cell phones are capable of storing a wealth of digitized information wholly unlike any physical object found within a closed container." https://en.wikipedia.org/wiki/Carpenter_v._United_States is also a good read, as is https://en.wikipedia.org/wiki/Riley_v._California :…

Those are all 4th amendment cases for warrantless searches violating privacy. The fifth amendment has very different criteria - with a proper warrant, your most private things are admissible evidence. You can't be compelled to testify but all your most private notes in a safe can be "interrogated"; in many situations you can't be compelled to testify against your spouse but any writings or recordings of what you said…

Thanks for the analysis. Something still sticks in my craw.

You can read a letter. You can play an audio recording. You can look at a photograph. You can print a word-processing document. Those feel like static records. Each tells a single story, no matter how many times it's read.

An ML model, on the other hand, is just a long sequence of floating-point numbers. There is no meaningful model "viewer" that spits out a text file or JPEG. The only thing a non-engineer human can do with an ML model is interact with it when it's running. Thus, any output is the product of both the human and the model. It doesn't feel like a record. It feels more like a performance -- and a performance by the interrogator, at that.

If an ML-model "record" is a special kind of record that needs manipulation to produce human-readable output, then it feels like we're back in the 5th amendment department. No, chatting with a chatbot is not torture. But it produces a kind of evidence that is an unwelcome collaboration between the interrogator and the record (unwelcome by the owner of the record). It's one thing to let a jury look at a screen full of numbers. It's another thing to say that an expert witness used those numbers in a chat session that printed "Yes, it was my owner, in the living room with the candlestick."

If you lock up a person and ask them the same question over and over again, eventually you'll get the answer you want. The same will probably turn out to be true for that person's chatbot. Should society allow that second case?

Re: I used GPT to build a search tool for my second brain note-taking system

#99

Post by tactiq »

alostpuppy wrote: post #45
senectus1 wrote: post #33

Earlier quoted context omitted.

MS is very close to this. they have chatGPT listening to meetings and taking notes then adding tasks to attendee calendars...

Link? Because I’m gonna need this.

https://tactiq.io/

Re: I used GPT to build a search tool for my second brain note-taking system

#100

Post by wazoox »

asdff wrote: post #77
dewey wrote: post #75

Earlier quoted context omitted.

I feel like this was the promise of voice assistants like Siri many years ago, turns out they currently still have issues turning on / off lights reliably.

I'll believe an AI model can update my calendar accurately only after it can correctly tell if I just said 15 or 50.

Just say "Meet Jane tomorrow at ten to four PM" and you're gold :)
Post reply on HN