omi_conversations: 06e7e97c-10b7-40c8-9d90-1d1268af984a
This data as json
| id | created_at | started_at | finished_at | title | overview | emoji | category | language | source | raw_json |
|---|---|---|---|---|---|---|---|---|---|---|
| 06e7e97c-10b7-40c8-9d90-1d1268af984a | 2026-04-07T20:24:35.674103Z | 2026-04-07T20:25:46.880725Z | 2026-04-07T20:28:29.487584Z | Exploring Personal LLM Knowledge Bases For Professionals | The user talks through an idea about personal knowledge bases powered by large language models (LLMs). They explain a recent concept circulating on Twitter where individuals collect materials on a topic (e.g., legal doctrine, case law, or general professional knowledge) and feed them into an LLM instead of just bookmarking or reading passively. The LLM then summarizes, organizes, and structures the information, continually refining its internal model so the user can later query it more effectively. The user wonders whether this approach would be professionally useful, particularly for someone at a law firm like Vic, and considers experimenting with Vic as a test case that might later be expanded to other attorneys or professionals if it proves valuable. | 💡 | technology | en | omi | {"id": "06e7e97c-10b7-40c8-9d90-1d1268af984a", "created_at": "2026-04-07T20:24:35.674103Z", "started_at": "2026-04-07T20:25:46.880725Z", "finished_at": "2026-04-07T20:28:29.487584Z", "structured": {"title": "Exploring Personal LLM Knowledge Bases For Professionals", "overview": "The user talks through an idea about personal knowledge bases powered by large language models (LLMs). They explain a recent concept circulating on Twitter where individuals collect materials on a topic (e.g., legal doctrine, case law, or general professional knowledge) and feed them into an LLM instead of just bookmarking or reading passively. The LLM then summarizes, organizes, and structures the information, continually refining its internal model so the user can later query it more effectively. The user wonders whether this approach would be professionally useful, particularly for someone at a law firm like Vic, and considers experimenting with Vic as a test case that might later be expanded to other attorneys or professionals if it proves valuable.", "emoji": "\ud83d\udca1", "category": "technology", "action_items": [], "events": []}, "language": "en", "source": "omi", "transcript_segments": [{"id": "bed74ab9-39cb-4eaf-b226-cebff4b9e294", "text": "Talking to me? A little bit. Bit. I have a work question for you. Would you benefit from a personal knowledge base? For work stuff?", "speaker_id": 0, "speaker_name": "User", "start": 5.912784217798617e-07, "end": 27.202288236618188}, {"id": "4ffab5f8-ca21-40c4-bb87-615ee156095c", "text": "What that would mean is you dump you start dumping things you have into, like, a database. And then an LLM organizes and cultivates it.", "speaker_id": 0, "speaker_name": "User", "start": 27.762288236618133, "end": 36.76228823661813}, {"id": "b18a7681-a4fb-45ad-8914-9f1c6e54dced", "text": "Then you can ask your LLM questions and it can you know what I mean? So it'd be like, legal doctrine, I don't know what case law. I I don't that's the question. Like, is that something that, like, professionally would be useful to you?", "speaker_id": 0, "speaker_name": "User", "start": 37.24228823661815, "end": 48.82228823661808}, {"id": "9db0c617-c62a-4a0e-9eb1-d34d0b87d084", "text": "Yeah.", "speaker_id": 0, "speaker_name": "User", "start": 49.7222932366185, "end": 50.282288236618115}, {"id": "e489bbd9-6e22-4ad7-b727-bf00c9b1a2b0", "text": "Yeah. More the excuse me, like, for, like, legal doctrine, you've already got tools that do this. Yeah. But for for help. Talking about just for me, are you thinking about stuff in general?", "speaker_id": 0, "speaker_name": "User", "start": 53.682288236618206, "end": 80.49387053489545}, {"id": "57ddaa38-ee99-4df8-809d-6956fe7830e9", "text": "Well, on Twitter for the last two weeks, everybody's been talking about this this guy came up with this scheme to create what he calls personal research knowledge bases.", "speaker_id": 0, "speaker_name": "User", "start": 81.1138705348958, "end": 89.9338705348955}, {"id": "ffd94941-b7cf-4e12-9bec-2e3059910a5d", "text": "It's like you've have a topic you're trying to learn about or whatever. And you start dumping it into an LLM instead of just reading stuff and book marking and what we normally do. Oh, okay. You start throwing it at an LLM, and the LLM both like, reads it, summarizes it, organizes it, right, structures it, So it's like an extra layer on top.", "speaker_id": 0, "speaker_name": "User", "start": 90.55387053489585, "end": 111.87386553489569}, {"id": "33e2d83a-96d6-41a9-957b-7c036a637b5f", "text": "And then you can start asking and it and it has this loop where it constantly kind of refines its own internal model of the of the information.", "speaker_id": 0, "speaker_name": "User", "start": 112.59387053489581, "end": 119.88387053489578}, {"id": "ce6ad3f7-dced-40f3-94e3-2095e7b9e7ef", "text": "And and and people are really excited about this. Oh, this is such a good idea. This is so powerful. And so my first thought is, like, can I do something with this?", "speaker_id": 0, "speaker_name": "User", "start": 120.51387053489543, "end": 129.47387053489547}, {"id": "f7e86658-0e9b-482b-b810-065b5005d1a9", "text": "And I'm like, well, I'm kinda doing my own thing. Like, I know how to use it for me. But then I was like, oh, maybe Vic maybe with his firm, this would be something. The idea would be like, like to play around with it with you, and if it's useful, then it would be like, oh, we can do this with people on a bigger scale.", "speaker_id": 0, "speaker_name": "User", "start": 129.99387053489545, "end": 148.94387053489572}, {"id": "4bea68ab-791d-41bc-8851-6c51e746edcb", "text": "Yeah. Maybe other attorneys, maybe just other professionals in general. But but I need, like, a anyway, that's that's the motive for the question.", "speaker_id": 0, "speaker_name": "User", "start": 149.55387053489585, "end": 157.39388053489574}], "geolocation": null} |