Transcript detail
Loading...
Public transcript context with linked callsigns, related nets, and analysis metadata.
Transcript
Public transcript text
Yeah, hey Kirk, but it's from the 8th Call District, Michigan. Even though I'm in Lacey, Washington, KVA, BNY. I was, I got munching on a cake pop from Starbucks here, so excuse me. One of the guys checked in and mentioned the server and it sounds like he has an LLM that's actually learning from voice. That's fascinating. I wanted to make that comment and when you guys talk about that stuff, that's extremely interesting stuff. The question I had and you may or the other gentleman may know or none of you may know, but it seems that there's all these competing large learning modules and I've often wondered why not tie two or three or more of them together and pool their data because it's almost like they're kind of reinventing the wheel like this one's learning and the one over somewhere else is learning something. They may be learning the same thing where if they were tied together they could pool that data. I don't know if anyone's ever thought of that, but I I've often thought of that when I hear, oh they're competing here and they're competing there. It's like they could pool this data, and they'd have a massive LLM and who knows maybe they could reach single area. I don't know. That was my question. I had another question too. Let me do this. You mentioned that SD card degrades over time. Is there a mode of storage that theoretically could last like you know like in a time capsule for 50 or 100 years where you could pull it out and it would still actually work without going 100 years in the future, but as far as what we know about its durability is that possible? It's based on what we know now. Anyway, those are my two questions, and I'll pass it back to you Kirk. Everyone have a good afternoon. I just got back from doing errands and hitting Starbucks, and it's beautiful out there. So we're going to head back out here pretty soon. KBAPNY.
Explore