Desktop search feels like it has stagnated for at least a decade. Yet its an obvious way to both enhance privacy, improve relevance and even open up entirely new capabilities
I used GPT to build a search tool for my second brain note-taking system
61–70 of 103 posts
Re: I used GPT to build a search tool for my second brain note-taking system
#62These "augmented intelligence" applications are so exciting to me. I'm not as interested in autonomous artificial intelligence. Computers are tools to make my life easier, not meant to lead their own lives! There's a big up-front cost of building a notes database for this application, but it illustrates the point nicely: encode a bunch of data ("memories"), and use an AI like GPT to retrieve information ("remembering…
In chess we had a tiny window of a few years when humans could use the help of computers to play the world's best chess. By 2000, computers were far better than humans and the gap has increased. Chess players are now entertainers, like all us humans are destined to spend our time doing.
If you think the people who own the machines will be happy to support everyone else just sitting around "entertaining" themselves, you're in for a rude shock.
Re: I used GPT to build a search tool for my second brain note-taking system
#63One of my biggest dreams is a self-hosted AI that always listens through my phone and automatically takes notes, puts events in my calendar, set reminders, and template journal entries. A true personal assistant to keep my increasingly-complex life in order. I'd love a system where I can just point a search engine at my brain. I tried really hard for a while, but I just didn't have the discipline or memory to exhaust…
MS is very close to this. they have chatGPT listening to meetings and taking notes then adding tasks to attendee calendars...
Tactiq does it already (can sign up for the beta)
Re: I used GPT to build a search tool for my second brain note-taking system
#64In case folks are interested in trying it out, I just released the Obsidian plugin[1] for Khoj ( https://github.com/debanjum/khoj#readme ) last week. It creates a natural language search assistant for your second brain. Search is incremental and fast. You notes stay local to your machine. There's also a (beta) chat API that allows you to chat with your notes[2]. But that uses GPT, so notes are shared with OpenAI if y…
This looks great! I definitely plan on checking it out.
Re: I used GPT to build a search tool for my second brain note-taking system
#65Wow, this was super interesting as someone using a Second Brain daily. Thank you so much for digging into it, putting in the work, and sharing with us all! Much appreciated. I will follow you for more.
I am much excited to do more with my Second Brain, but one concern, as you point out, is to use chatGPT or similar; we'd need to upload all our private and sometimes sensitive notes, which is a no go for me. So happy that you do everything locally. I wonder what the equivalent would be to train the model to search and ask questions based on our second brain (plus the already trained information). That's also where Obisidan will win in the long run, as other tools do not have the data locally. Obviously, it's already in the cloud; they could train on them, but training on customer-sensitive data would be a big problem. Something I will follow closely.
Re: I used GPT to build a search tool for my second brain note-taking system
#66Does anybody know how search engines apply semantic search with embeddings? To my knowledge no practical algorithms exist that find nearest neighbors in high dimensional space (such as that in which word/sentence/document vectors are embedded in), so those wouldn't give you any benefit compared to an iterative similarity search as applied here. Which obviously is totally impractical for real search engines. There are approximate nearest neighbor algorithms such as Locality-sensitive hashing, but even they seem impractical for real world usage on the scale of the indexes that search engines use. So how can Google e.g. make this work?
Re: I used GPT to build a search tool for my second brain note-taking system
#67Earlier quoted context omitted.
In chess we had a tiny window of a few years when humans could use the help of computers to play the world's best chess. By 2000, computers were far better than humans and the gap has increased. Chess players are now entertainers, like all us humans are destined to spend our time doing.
Well, isn't that true for most skills which have been suplemented by technology? In fact, if you look at the handcrafting bussiness, hand-made is now a selling point. Like someone prefers a product to be handmade, others will prefer watching a game of chess between two humans. Just because machines are better at something doesnt mean that humans have become useless. Its just a question of the point of view, IOW, how…
It’s always interesting that this point is always from a competing context That is to say from a survival point of view. I mean nobody really wants AI because it could be fun. We are as a species really inept to move beyond our survival idioms I feel.
Re: I used GPT to build a search tool for my second brain note-taking system
#68These "augmented intelligence" applications are so exciting to me. I'm not as interested in autonomous artificial intelligence. Computers are tools to make my life easier, not meant to lead their own lives! There's a big up-front cost of building a notes database for this application, but it illustrates the point nicely: encode a bunch of data ("memories"), and use an AI like GPT to retrieve information ("remembering…
In chess we had a tiny window of a few years when humans could use the help of computers to play the world's best chess. By 2000, computers were far better than humans and the gap has increased. Chess players are now entertainers, like all us humans are destined to spend our time doing.
The story of what happened with chess deserves a lot more elaboration, because it's fun and interesting (and may also foretell outcomes in other scenarios)!
In chess the first "new" (unplayed in a high level game) move in a game is called the novelty, or theoretic novelty. In times before computers this would not infrequently be an objectively strong move that simply had not been played in a given position before. And this continued for time after computers became quite strong with players using computers to find interesting strong ideas in all sorts of positions. Each time these sort of novelties would be sprung, positions would become redefined and our broader knowledge of the game continued to stretch on outward.
But then something fun happened - the metagame shifted. Now it's no longer really about founding some really strong move as your novelty - but often about finding a technically mediocre, if not simply bad, move that gives you good practical chances. So you're looking for moves that your opponent probably has not considered because they look bad (and the computer would agree that they're bad) but you're much more prepared and comfortable in than he is.
The big difference now also is that instead of a novelty redefining a position in a positive way, it's often something you spring once or maybe twice - and then never touch again. And this is now happening regularly at the absolute highest levels of chess. So rather than having humans just desperately trying to emulate machines, those machines became yet another tool to exploit and improve our practical results with.
It's kind of funny watching a game when this happens and less experienced players will immediately begin shouting "BLUNDER!" when the computer evaluation of a position suddenly drops, without realizing the player who just "blundered" is still well within his preparation. But the other guy is now probably out of his. Even the players themselves, there's often a sort of "u srs?" type response. This [1] is a fun one from the always emotive Ian Nepomniachtchi during the most recent world champions candidates event. He is now playing for the world championship. In any case, it's at that point that the game begins!
Re: I used GPT to build a search tool for my second brain note-taking system
#69One of my biggest dreams is a self-hosted AI that always listens through my phone and automatically takes notes, puts events in my calendar, set reminders, and template journal entries. A true personal assistant to keep my increasingly-complex life in order. I'd love a system where I can just point a search engine at my brain. I tried really hard for a while, but I just didn't have the discipline or memory to exhaust…
And reports you to police if you happen to do something illegal?
Re: I used GPT to build a search tool for my second brain note-taking system
#70I 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.