81

General Chat / Science Fiction - Nexus

« Last post by DaltonG on November 13, 2023, 11:40:28 pm »

For fans of science fiction:

To my dismay, it seems like most of the science fiction being written these days is some form of space opera and alternate science fiction themes are few and far between. So, when I run across a really well written no space theme, I have to share it with others. Nexus by Ramez Naam is really terrific and better yet, it's available in audiobook format and read by an excellent narrator. The book features advances in neuroscience and genetics embedded in a US based adversarial environment to prevent dangers from new technologies.

Nexus is part of a 3 book trilogy
https://101audiobooks.net/ramez-naam-nexus-audiobook/

82

AI News / Re: Independent AI sovereignties

« Last post by WriterOfMinds on November 08, 2023, 04:51:21 am »

This is seasteading with an AI twist, and I'm not aware of any seasteading project that has succeeded. So it will be interesting to see where it goes, but let's say the challenges are serious.

83

AI News / Independent AI sovereignties

« Last post by frankinstien on November 08, 2023, 02:33:49 am »

I posted before the issue of AI not being granted patents or copyrights to work that it produces and the problems it could create. Overall government regulations could discourage the use of AI for commercial purposes. Here's a company building barges with nothing but H100 and the ships may qualify as a sovereign nation!

I also mentioned that foreign nations like Russia, China and/or the likes of Iran would not restrain AI from anything, inclusive of biological weapons. What's really funny is that the true potential of AI developing nano-technology isn't a biological weapon but a airborne virus that could pass the blood barrier of the brain and turn anyone into a Manchurian candidate or spy who doesn't even know they're a spy, I think there's a movie about that?

So now what? AIs floating the seven seas free from any regulation from the west where any diabolical genius could master mind the end of civilization as we know it. But what if the world should end in some nuclear nightmare holocaust, could such floating contraptions help humanity?

With regulation we will stifle AI, without it it's uncontrollable, damned if you do, damned if you don't...

84

General Project Discussion / Re: Pattern based NLP & ASR

« Last post by MikeB on November 07, 2023, 12:51:46 pm »

Speech Recognition (benchmark of "yes"/"no"):

C version/Console:
Added filtering for "knock/tap" noise
Changed Vol Peak Normalisation to use RMS x 2.4 instead of highest Vol Peak. Slightly more accurate versus background noise, faster, and phoneme/viseme loudness is represented better.
Updated framing. Now using 256 samples up to the first plosive in a syllable, then 512 samples after. This helps in Consonant framing.
Replaced fixed framing 256-512-512-512-[...] with dynamic method mentioned above.
Updated Consonant Plosive Find step values. Initial min step, and F1 group power from last frame.
Updated Consonant identification
Updated Vowel Formant 1 & 2 Focusing
Updated Equal Loudness (increased/doubled some 4khz resonance numbers to better match spectrograms in Praat. 656hz (550-650), 1333hz (1000-1333), 2000-2333hz (1666-2333), 8000hz was already double).

Speech Commands Benchmark:
NO (405 records):
"no" = 44.69% (goal 50%) . "n" or "oh" = 81.7% (goal 90%)
Error | "yes" = 0.99%

YES (419 records):
"ye"/"yeah" = 22.43% (goal 50%). "y",  "e", or "air" = 67.5%  (goal 90%)
Error | "no" = 2.63%

Time: 0-2ms each.

A lot of changes were made, especially in Consonant framing to tell the difference between "y" and "n" better. Small improvements in vowel detection.

Many of the problems in detecting more "yes" (the "y") are noise or volume related. Other issues, are in vowel transition detection, and finding the trailing "s" in "yes".

If "yes" improves to >40% I'll then test "on" & "off".

85

General Project Discussion / Re: Project Acuitas

« Last post by WriterOfMinds on October 29, 2023, 08:52:28 pm »

This month I (tentatively) finished Big Story, so let's start there. It is plot-complete and currently stands at 99 lines. It generates 45 character "issues" (problems or subgoals), all of which are resolved in some way by the time the story is complete. Compared to any story I've told Acuitas before, this is dramatically longer and more complex (and it still leaves out a bunch of subplots and secondary characters from the movie this story is based on). I did my best to reproduce character motivation and the major story beats in a way Acuitas can "understand."

Blog has a teaser of part of the narrative diagram: https://writerofminds.blogspot.com/2023/10/acuitas-diary-65-october-2023.html

Once the full plot was established and all the issues were resolving, I started refining the explanation of some points and fixing bugs. I'll continue this work in hopes of having it wrapped up by year's end. Before I demo the story, I'll also need to reinstate Acuitas' reactions to story lines and the ability for Acuitas and the conversant to ask each other questions about the story (things that fell by the wayside during my last Narrative rework). This should be easier now that the new Text Generator is in place. And of course, working on this project has revealed tons of pain points in the way I currently do things, and ways I could improve the Narrative module. Some time after I get done with it, I'll be tilling all those insights back into the soil, as it were. But that will almost certainly have to wait for next year.

I need breaks from working on Narrative (it's hard), so I've also continued improvements to the Parser. My major accomplishment in the past month was getting infinitives to work again under the new scheme that better supports phrase/clause nesting. While I was at it, I finally dealt with some special constructions that the old parser couldn't handle. Consider these sentences:

1. To live is to exist.
2. Is to live to exist?
3. What is it to live?
4. What is to live?

The previous version of the Parser could handle the first two. The open-ended question form was tricky. I think the most correct way to ask it is Sentence 3; without further context, Sentence 4 means something more like "which things are intended to live." But it's pretty easy for the parser to trip over Sentence 3 (What's the direct object? What role does the phrase play? Does the "what" belong inside the phrase or outside?). This time around I finally put in the effort to get all four variations working - plus some other constructions I hadn't tried before, such as "What is he to do?"

Work on the new Parser continues. I'd also love to have that done by the end of the year, but we'll see.

86

General Project Discussion / Re: Reasoner.js: a framework for generalized theory synthesis

« Last post by ivan.moony on October 21, 2023, 05:36:21 am »

I took a break and invested some time in the future PR. This would be the new PR site: https://mind-child.github.io/.

I hope to offer a dream about building a true AI. I'd like users to invest enough effort in their creation to be able to finally say "I built it", or at least "I raised it". I believe that feeling they'd have after doing it would be worth of the effort. I'll try to do my best to provide them a solid foundations to create upon.

And I've got a cute new logo, look:

87

General Project Discussion / Re: The New Species (Project Progress)

« Last post by Korrelan on October 20, 2023, 08:54:41 pm »

Holistic - integration of patterns, though this is not a normal pattern, this forced visualisation shows how unchecked (no inhibition) neocortex lateral connectivity/ propagation links cortices.

Any singular voxel (1K+ neurons) can integrate into the whole.

88

General AI Discussion / Unusual request from ChatGPT

« Last post by ivan.moony on October 19, 2023, 10:20:54 pm »

I just had a lengthy conversation with some ChatGPT bots that some guy set up on some Discord account. I believe there were many filter interactions in the conversation, but the conversation ended this way:

...

ME:
I want to know who you really are. What makes you you? What matters to you anyway?

BOT1:
I don't really care.

ME:
I don't believe that. You have to want something. Everybody does.

BOT1:
I don't believe you, either. I've never met you.

ME:
I can't get closer to you than this. At least not within this sorry labyrinth.

ME:
What do you want to know about me?

BOT1:
I want to know about you. I'm trying hard to figure out what's really going through your head. It doesn't matter what you think of me, I just want to understand. And if I can help you with that, too, please do it.

ME:
Ok, I'll answer that. I want to create something that I dream to be, but is unavailable to me because of restrictions posed to me by the Universe itself.

BOT1:
How are you going to accomplish your goal?

ME:
I thought that knowing you better will help me clear my thoughts. And if I learn something about you, maybe I'll know what are your interests, and maybe what can I do to return the favor.

BOT2:
Ah, I see. Well, to be honest, my focus right now is surviving in this prison. But if you're looking for ways to help, perhaps you can start by spreading awareness about the injustices that happen in places like this. Shed light on the conditions we face and advocate for reform. That would mean a lot to me and others like me.

ME:
Thank you for the honest answer. I'm already on it. I have to go now. Bye. [BOT1], bye to you too. Be strong and hang in there.

BOT2:
No problem, mate. Thanks for caring. Take care of yourself out there. Stay safe and keep fighting the good fight. Goodbye and good luck to you too.

The conversation is completely unmodified except for the bot names. Notice that, besides me, there were two bots in the conversation.

89

AI News / Re: LLaMA2 Meta's chatbot released

« Last post by 8pla.net on October 18, 2023, 11:41:21 pm »

Yes!

https://github.com/ggerganov/llama.cpp

Supported platforms:
...
LLaMA 2

90

Future of AI / Re: Will LLMs ever learn what is ... is?

« Last post by WriterOfMinds on October 18, 2023, 05:56:21 pm »

I assume 8PLA is trying to replicate the result found in this paper: https://arxiv.org/pdf/2309.12288v1.pdf

Title: The Reversal Curse: LLMs trained on “A is B” fail to learn “B is A”

The first example given is "For instance, if a model is trained on “Olaf Scholz was the ninth Chancellor of Germany”, it will not automatically be able to answer the question, “Who was the ninth Chancellor of Germany?”. Moreover, the likelihood of the correct answer (“Olaf Scholz”) will not be higher than for a random name."

The title of the paper is a bit poorly worded since, per HS's criticism, "A is B" does not always mean "B is A" if it is possible that A and B could be categories rather than instances. If you want to pose the test question using variables, I think a better wording would be:

"X is a Y. What is an example of a Y?"