Searching...
14 results for “memory performance”
...memory? Like, can it be improved? Yep. It can be improved. And there are now some really good brain training apps as well that can help to get people to work better with attention, to have NVAC tests, so remember information a couple of information p
2:19:25 / 2:20:31
from transcript2:19:25 – 2:20:31
...inside of MCP. When they launched, with the initial set of servers, they had a memory server in there, which I I would recommend as, like, that's where you start with memory. But I think, like, if there was, a better, memory abstraction, then a lot o
14:56 / 16:01
from transcript14:56 – 16:01
...memory of my preference. It's more about what explicitly said. And I'm trying to figure out at what level that gets solved. You know, like, is it, do these memory products, like the MGPPs of the world, create a better way to implicitly extract prefer
1:00:45 / 1:01:47
from transcript1:00:45 – 1:01:47
...of memory. Then go back to my previous point. Right? The long context window basically means, like, when you're when the the things that you are processing longer and longer, right, like, how can you how can you have enough memory to process those. R
18:43 / 20:21
from transcript18:43 – 20:21
...bad memory for names. All of these mnemonic devices that have been studied over the last hundreds of years mostly rely on creating elaborate pictures in your mind. So, like, memory champions, people who do competitive remembering, will tell you that
2:21:43 / 2:22:47
from transcript2:21:43 – 2:22:47
...of you to ship to everyone else. Simple answer is there's no standardization for a while because everyone will experiment with their own stuff. And I think entropic success of MCP suggests that basically no one else but the big labs can do it because
1:04:29 / 1:05:32
from transcript1:04:29 – 1:05:32
“TikTok scrolling destroys your memory worse than random guessing”
...TikTok, memory performance is barely better than random guessing. So, apparently, if you What does that mean? So, basically, if you opened TikTok and scroll through a lot of lot of, sort of clips Uh-h
9:31 / 10:00
market insight9:31 – 10:00
...And, like, when you have a bug, is it the memory or is it something core?
19:41 / 20:45
from transcript19:41 – 20:45
...to memory very often. SRAM, on the other hand, is unbelievably fast, but has low capacity. And so we wanted to use SRAM, but if you build a normal sized chip, you can't hold a model. And so by going to wafer scale, we were able to put down a huge amo
9:00 / 10:05
from transcript9:00 – 10:05
...memory, and there are three projects out there, MemZero being one of them. But the thing that's interesting for me, right, is and so we see this in ChattyPete, other things right now where it does have the the notion of a longer term memory can pull
1:10:28 / 1:11:31
from transcript1:10:28 – 1:11:31
...Then go back to my previous point. Right? The long context window basically means, like, when you're when the the things that you are processing longer and longer, right, like, how can you how can you have enough memory to process those. Right? And,
19:25 / 20:29
from transcript19:25 – 20:29
...memory retrieval gone wrong. He didn't actually want that. So even in a product like ChatGPT that spent a lot of time on memory, it's nontrivial. And I think my take is the writing of memories is tricky. Like, when actually should the system write me
38:16 / 39:19
from transcript38:16 – 39:19
...is this special memory? I don't know if you can talk about it. You mentioned a minute plus, but is there some sort of, like, measure that you have? Is it, like, can you keep it for half an hour, or what is the limit on that? There is no, like, fundam
11:13 / 12:20
from transcript11:13 – 12:20
...memory is definitely, like, a feature that has been under under invested in, by the field. But I would say that it is kind of difficult to invest in memory in this very, like, task centric regime. Mhmm. Because if you have, like, a a bunch of these,
31:35 / 32:39
from transcript31:35 – 32:39
Have a podcast?
Run your podcast through Clypt
Get ranked clips, hooks, and ready-to-post copy from your own episodes. Free to try.