ffwd

7 ms·

1.
by ffwd 1mo ago
LLMs are still missing a part of working memory. Part of working memory is being able to attend to small amounts of information and then understand and parse all the pieces of that information. When LLMs use their "working memory"
2.
LLMs might not learn concepts(pastebin.com)
4 points by ffwd 2mo ago|0 comments
3.
by ffwd 2mo ago
IMO consciousness is different from working memory, at least to a certain degree. The inner mind (working memory) is different from consciousness from sensory stimuli. You can stand outside and take in the environment and have a relatively
4.
by ffwd 3mo ago
Just some speculation, but, I think humans have on the one hand a lot of degrees of freedom in behaviors and thoughts they can do, but at the same time all that freedom is reigned in by our biological needs, like preserving the integrity of
5.
by ffwd 3mo ago
I just want to add that the "recursive" part of recursive self improvement is by no means a given, even if an AI can improve itself. Recursive self improvement is by its nature a step wise behavior not a continuous one, I would ar
6.
by ffwd 4mo ago
Well in that case I'm not sure where you're going. I agree that hamsters probably have a similar consciousness to ours, which is kind of the point I was trying to make. I think that consciousness comes before self-awareness, even
7.
by ffwd 4mo ago
I'm not sure consciousness and self-awareness are the same thing. First is we can be conscious when we sleep/during REM sleep, where it's arguable we are not self-aware. And if not that, we can even do it when awake, for exam
8.
by ffwd 4mo ago
Personally I'm not a fan of the emergence story, for a number of reasons. First is, it doesn't really make sense for consciousness to have emerged gradually through natural selection. When we and animals are conscious, the whole b
9.
by ffwd 6mo ago
Sorry I messed up how I wrote that. I meant to only remove metadata columns, not all columns.
10.
by ffwd 6mo ago
I find that this happens when you enter folders that have media files like audio files, video files and so on. One way to fix it is to enter one such folder, then remove all columns (like file name, date modified - those columns) and remove
11.
by ffwd 6mo ago
I feel like I kind of borked the last paragraph so I want to clarify something. The point is basically that since these repeating patterns are different every time, they are not emergent. They don't really "exist" except as m
12.
Emergence Isn't Real(pastebin.com)
3 points by ffwd 6mo ago|2 comments
13.
by ffwd 6mo ago
Sorry for pastebin but it's all I can do atm! Would appreciate your views
14.
by ffwd 6mo ago
Would appreciate your views on this!
15.
by ffwd 8mo ago
Maybe it's looked down upon to complain about downvotes but I have to say I'm a little disappointed that there is a downvote with no accompanying post to explain that vote, especially to a post that is factually correct and nothin
16.
by ffwd 8mo ago
Even though I think it's true that it's lossy, I think there is more going on in an LLM neural net. Namely that when it uses tokens to produce output, you essentially split the text into millions or billions of chunks, each with p
17.
by ffwd 10mo ago
LLMs _can_ think top-to-bottom but only if you make them think about concrete symbol based problems. Like this one: https://chatgpt.com/s/t_692d55a38e2c8191a942ef2689eb4f5a The prompt I used was "write out the cha
18.
by ffwd 11mo ago
That's a good point and I agree. I'm not a neuroscientist but from what I understand the brain has an associative memory so most likely those patterns we create are associatively connected in the brain. But I think there is a diff
19.
by ffwd 11mo ago
I think something that's missing from AI is the ability humans have to combine and think about ANY sequence of patterns as much as we want. A simple example is say I think about a sequence of "banana - car - dog - house". I c
20.
by ffwd 1y ago
Yeah interesting I have to think about this.
21.
AI might not recursively self improve (part 2)(secondsight.dev)
4 points by ffwd 1y ago|3 comments
22.
Recursively self improving AI might not work without these functions(secondsight.dev)
7 points by ffwd 1y ago|0 comments
23.
by ffwd 1y ago
No, the new algorithms used to be determine this was created by ICM-CSIC who are also the publishers of this article. Also the authors of the paper is involved with the article, there is for example this quote: “We are witnessing a true rev
24.
The universe is not based on information(secondsight.dev)
1 points by ffwd 1y ago|0 comments
25.
by ffwd 1y ago
Yep definitely agree with this.
26.
by ffwd 1y ago
Why can't it be algorithmic? If the brain uses the same process on all information, then that is an algorithmic process. There is some evidence that it does do the same process to do things like consolidating information, processing th
27.
by ffwd 1y ago
I think humans have some kind of algorithm for deciding what's true and consolidating information. What that is I don't know.
28.
by ffwd 1y ago
This to me is the paradox of modern LLMs, in that it doesn't represent the underlying domain directly, but it can represent whatever information can be presented in text. So it does represent _some_ information but it is not always cle
29.
by ffwd 1y ago
But the problem is we don't know how it works. It's not about assuming consciousness is outside of physical reality or something like this, it's simply the fact that we don't have an understanding of it. For example if w
30.
by ffwd 1y ago
This could be true but it depends on the person. Marc Andreesen is a billionaire and he wears Rolexes quite prominently. In fact a lot of rich people wear Rolexes lately.

More ›