Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon.

Pages: 1-4041-

WAIFU AI SOFTWARE

Name: Anonymous 2013-09-30 2:19

What?
A simple chatbot isn't enough, nor is a barcode kanojo.
We need something that can pass for human, and if not human, can at least pass for sentient. We need something that thinks for itself, something that learns.
Why?
I believe AI is the only thing holding us back from a sudden singularity.
I believe sometimes a 'waifu' just doesn't cut it.
I believe we CAN make AI, not by trying, by doing.
When?
Now.
Who?
You, me, we. We can do it!
How?
I don't know yet, but I'll be damned if I don't find out!

This is a challenge.

Name: Anonymous 2013-09-30 2:23

AI is already solved. Check out Mindforth for more info.

Name: Anonymous 2013-09-30 8:25

>How?
>how are neural network formed?

Name: Anonymous 2013-09-30 9:21

http://www.mitsuku.com/

You need never feel lonely again! Mitsuku is your new virtual friend and is here 24 hours a day just to talk to you. She
learns by experience, so the more people talk to her, the smarter she becomes.

Name: Anonymous 2013-09-30 9:46

>>4
AI my ass.

Human: Why are you bullying me?

Mitsuku: Have you told someone in authority? They may be able to help you. I would advise that you let someone know before it gets out of hand Aaaaa.

Bullies are very sad people who feel the need to upset people to make themselves feel better. The are often very lonely people really and only do it for attention.

Even so, you shouldn't worry about it. You should tell someone about it as soon as possible. 

Name: Anonymous 2013-09-30 20:45

>>2
>AI is already solved
u wot m8

Name: Anonymous 2013-10-01 10:45

>>2
MindForth
Googled this. The rants on the site, and the way they structured their code, makes my skin crawl. This is a travesty against the workings of the human brain.

Name: Anonymous 2013-10-01 11:04

>>7
I'll have you know that Arthur T. ``Mentifex'' Murry is a very accomplished /prog/rider who has provided thousands of insightful posts to this board. Back to fucking /g/ with you, you know-nothing idiot.

Name: Anonymous 2013-10-01 13:25

>>8
back to pedorider

Name: Anonymous 2013-10-02 11:37

>>8
>surprise box
That Mindforth shit is half assed. It has not "solved AI", but I will be able to glean a few things from it.

Name: Anonymous 2013-10-02 12:26

>>10
You haven't even read AI4U.

Name: Anonymous 2013-10-02 14:06

Name: Anonymous 2013-10-02 15:23

install gentoo

Name: Anonymous 2013-10-02 20:42

>>10
This published book claims otherwise: http://www.amazon.com/The-Art-Meme-ebook/dp/B007ZI66FS

How many books do you have published?

Name: Stephan Hawking 2013-10-02 23:15

>>14
How many have YOU published?

Name: Anonymous 2013-10-02 23:50

>>15
Hundreds. I ripped ASSTR, ran it all through a drunk markov chain generator, divided the output into 200 page sections, and published it on the Kindle. Thanks Amazon!

Name: Anonymous 2013-10-03 2:07

>>16
How much you get from it?

Name: Anonymous 2013-10-03 2:14

>>17
Three whole dollars and ninety nine cents! (Minus Amazon's cut.)

Name: Anonymous 2013-10-03 2:14

>>17
Start be going to loli threads on /b/

Name: Anonymous 2013-10-04 9:54

check 'em

Name: ☭Anonymous☭ 2013-10-05 0:27

>>20
Nice try.

Name: Anonymous 2013-10-05 0:40

Name: Anonymous 2013-10-05 2:13

AI is hard yo

Name: Anonymous 2013-10-05 3:48

noone said life was going to be easy? =)
but how's it going?

Name: Anonymous 2013-10-05 4:07

>>22
nice dubs bro

Name: Anonymous 2013-10-05 4:10

>>24
Waiting to order a new computer, this one doesn't really have the specs to do what I'm planning as a prototype!

Name: Anonymous 2013-10-05 7:03

>>26
hmm..
better algorithms > faster machines

Name: Anonymous 2013-10-05 8:45

>>27
I only have 512 megabytes of RAM and a Pentium M!

Name: Anonymous 2013-10-05 10:05

progriders.org

Name: Anonymous 2013-10-05 10:22

>>29
Please don't spam my thread, thanks.

Name: Anonymous 2013-10-05 10:25

>>30
I'm sorry, I was only trying to have a little fun. I won't do it again, promise. Can you forgive me?

Name: Anonymous 2013-10-05 10:28

gnu.org

Name: Anonymous 2013-10-05 10:40

dubs.org

Name: Anonymous 2013-10-05 11:20

autism.org

Name: Anonymous 2013-10-06 15:07

Neural networking.

Name: Anonymous 2013-10-06 15:45

I would write the best AI ever if I only could understand how universal generalization works and how the hell is it a valid inference rule. Here's the Wikipedia article: https://en.wikipedia.org/wiki/Universal_generalization

Supposedly you can generalize if the function of some dataset works for all x and the function has no free variables, but a function that checks the MD5 hash of its input against a checklist has no free variables yet you can't generalize from that! Apparently this is because said checklist doesn't have ''every possible'' input, but if I had every input then I wouldn't need to generalize.

So, what the fart is universal generalization? I have no clue at all! Help a retarded niggerfaggot out!

Name: Anonymous 2013-10-06 15:48

>>36
Oh wait, MD5 has free variables. But I'm pretty sure I could come up with a hash algorithm that has no free variables and make the same argument... or could I? Is this the only problem?

Name: Anonymous 2013-10-06 16:06

>>37
Ok i did come up with an hash algorithmhm tha t has no free variables so im pretty sure I'm a retard and I don't understand generalization.

Why am I trying to understand it then? Because I figured you can't actually measure the generalization ability of something (at least not rigorously or unsupervised) so I'm trying to figure out what's the actual meaning of generalization and all this philosophy is fucking ripping my soul apart.

Name: 36,37,38 2013-10-06 16:11

>>36,37,38
Also I'm pretty sure, if not absolutely positive, that if the code of a function is smaller (information-wise) than the dataset it's trying to fit to, then that function generalizes perfectly. So I could start with something like a MD5 table and then work to make the function smaller and smaller but there's the exact same problem as with neural nets: convergence is shittier and slower than my shit-stained anus.

Name: Anonymous 2013-10-06 19:39

>>36,37,38,39
Intense.

Name: Anonymous 2013-10-06 19:42

>>40
What the fuck was the point of that shitty, useless reply? What could possibly have compelled you to type that out? Do you feel all tough and badass now that you've made such a cynical, apathetic comment? Perhaps /b/ would be better suited for you, since you can find lots of like-minded edgy teens from there.

Name: Anonymous 2013-10-06 19:48

>>41
LEL U MAD XD

Name: Anonymous 2013-10-06 19:52

>>41
watch out we got a badass over here

Name: Anonymous 2013-10-06 19:53

>>36
I tried proving the wikipedia example here and got this
http://imgur.com/9FUPMlD
I don't think I've used universal generalization though

Name: Anonymous 2013-10-06 20:30

>>41
Eh, to bump the thread. I ain't even cynical apathetic or anything.

Name: Anonymous 2013-10-06 20:45

>>28
fair enough... so what are you planning to do, if i may ask?

Name: Anonymous 2013-10-06 22:00

>>46
Well, I'll look over Mindforth once more and do research in the field.

I was thinking of trying a neural networking approach first.

I still need to decide what language to use.

Name: Anonymous 2013-10-07 10:49

>>47
Well, I'll look over Mindforth
lol wut?

Name: Anonymous 2013-10-07 12:48

>>48
look over (third-person singular simple present looks over, present participle looking over, simple past and past participle looked over)

    (transitive) To give a brief visual inspection.
    (transitive) To scan-read and check for errors.

Name: Anonymous 2013-10-07 13:06

>>48
>le pedophile sage

Name: Anonymous 2013-10-09 22:43

>>50
>shitty forced meme

Name: Anonymous 2013-10-10 0:45

>>51
>le pedophile criticism

Name: Anonymous 2013-10-16 3:52

pedophile bump

Don't change these.
Name: Email:
Entire Thread Thread List