First  Prev  1  2  3  Next  Last
The AI
6080 cr points
Send Message: Send PM GB Post
32 / M
Offline
Posted 7/8/17 , edited 7/8/17
Guys, I am here to present an idea about how I would create an AI.

For those that are familiar with text mining and word cloud, you associate words based on the article and frequency.

In my case, you search Google for all the articles and texts you can find and store it into the database. Then you pull them up one by one and you chop it up by each sentence and article. So that each sentence or article represents a general idea.

For instance, I will type in the words "water" and "thirsty", if parse the sentence correctly and based on words frequency count. I would probably get "drink" 98%, "sip" 2%, etc. Based on the frequency count of the words. And that my friend, is your AI.

I can also type drink, thirsty, and probably find something like aqua cola.

But anyway, it is all about text mining, enjoy


P.S Credited to my previous Python boss
runec 
41006 cr points
Send Message: Send PM GB Post
Offline
Posted 7/8/17 , edited 7/8/17
Er, that's not an AI. That's just data parsing.

6080 cr points
Send Message: Send PM GB Post
32 / M
Offline
Posted 7/8/17 , edited 7/8/17

runec wrote:

Er, that's not an AI. That's just data parsing.



Close enough, you put in any words and it gives you an answer based on frequency count.


P.S Haven't worked out the multiple words search detail though, but you get the idea, I'd go with n-gram
Posted 7/8/17 , edited 7/8/17
We don't need another Siri.
6080 cr points
Send Message: Send PM GB Post
32 / M
Offline
Posted 7/8/17 , edited 7/8/17

LingLingJuju wrote:

We don't need another Siri.


This is more like a Mc Guyver type of thing, like for instance if you type in "snake" "bitten" it would give you a solution
runec 
41006 cr points
Send Message: Send PM GB Post
Offline
Posted 7/8/17 , edited 7/8/17

fredreload wrote:
Close enough, you put in any words and it gives you an answer based on frequency count.


No offense but its not even close to AI. >.>

What you're describing is pretty standard data processing. Anyone whose done any sort of coding has used and/or touched upon the same principles.
6080 cr points
Send Message: Send PM GB Post
32 / M
Offline
Posted 7/8/17 , edited 7/8/17

runec wrote:


fredreload wrote:
Close enough, you put in any words and it gives you an answer based on frequency count.


No offense but its not even close to AI. >.>

What you're describing is pretty standard data processing. Anyone whose done any sort of coding has used and/or touched upon the same principles.


Then define a good AI for me, maybe I've missed the target
runec 
41006 cr points
Send Message: Send PM GB Post
Offline
Posted 7/8/17 , edited 7/8/17

fredreload wrote:
Then define a good AI for me, maybe I've missed the target


I think you might be misinterpreting the term AI somehow? Or not using the true definition? I mean, achieving actual artificial intelligence would be a huge technological breakthrough. Even in a more game/programming sense though what you're describing isn't really what I'd call AI. I guess if you were just using the term AI in a game programming sense for example you might be closer to the mark. But actual AI, no.

What you're describing is just data parsing. The program isn't making observations or decisions. It's not exhibiting any sort of cognitive functions. It's just processing data and providing a result. You say "drink", it parses and returns "sip". It doesn't understand what "drink" or "sip" actually are or what you actually intended by asking. It's just comparing your data entry to other data and returning a likely association.

Chat bots have been doing that sort of thing for ages.



Posted 7/8/17 , edited 7/8/17
kizuna ai is the only ai I need in my life.
42533 cr points
Send Message: Send PM GB Post
35 / M
Offline
Posted 7/8/17 , edited 7/8/17


Haven't yet heard an explanation of AI that gets past the Chinese Room problem.

http://www.iep.utm.edu/chineser/
6080 cr points
Send Message: Send PM GB Post
32 / M
Offline
Posted 7/8/17 , edited 7/8/17

runec wrote:


fredreload wrote:
Then define a good AI for me, maybe I've missed the target


I think you might be misinterpreting the term AI somehow? Or not using the true definition? I mean, achieving actual artificial intelligence would be a huge technological breakthrough. Even in a more game/programming sense though what you're describing isn't really what I'd call AI. I guess if you were just using the term AI in a game programming sense for example you might be closer to the mark. But actual AI, no.

What you're describing is just data parsing. The program isn't making observations or decisions. It's not exhibiting any sort of cognitive functions. It's just processing data and providing a result. You say "drink", it parses and returns "sip". It doesn't understand what "drink" or "sip" actually are or what you actually intended by asking. It's just comparing your data entry to other data and returning a likely association.

Chat bots have been doing that sort of thing for ages.





I understand your idea of AI and I also have searched around it for years . Now the AI you've mentioned would contain artificial consciousness. Anything with a consciousness would think for itself. But then it would die eventually. You could theoretically create a brain simulation in 3D, but then you will have to create force, which requires a 3D simulation of space time.

Now my idea of AI is one that brings the answer to you when you need. Think of Google search, a Google search bring the most likely searches of website to you. And there is the word cloud if you have done text mining before. Sort of like fill in the blanks for you. Each sentence is essentially an idea, check out n-gram sorting method. The validity of this idea is further reinforced by frequency count.

Alright here is an example.
I eat bread.
I eat a lot of bread.
I eat rice.
I eat rice.
I eat poo.

Doing a sorting by frequency I get:
bread 2 40%
rice 2 40%
poo 1 20%

Alright so most people eat bread or rice but not poo.

Now if I check the snake bitten:
suck the blood out 40%
pee on it 40%
antidote 20%

You could even include a time scale and snake type, this is where the power of text mining comes in.

Long story short. a conscious AI is hard to manage, and think like a human.
Text mining contains ideas with a vast database.

If you don't like my idea, feel free to give me a better one , personally I think this is a pretty cool idea, but right I've never participated in a trivia game
10072 cr points
Send Message: Send PM GB Post
Offline
Posted 7/8/17 , edited 7/8/17
If you can code you'll want to head over to Silicon Valley and rent an office. That's where all the easy VC money for "AI" is at the moment. You need some marketing buzzwords and a slick presentation. You're only problem then will be working out how to keep as much of the VC money (and generate a few more rounds) while you build a pretty front end for your data mining algorithm.

How's that for a better idea?

We can't even define what human intelligence is (and I'm sure we all know some people without any), so creating it in an inorganic form is well off the cards.
5558 cr points
Send Message: Send PM GB Post
25
Offline
Posted 7/8/17 , edited 7/8/17
AHHHHH another Fredthread :D


I eat poo.


hahaha
27031 cr points
Send Message: Send PM GB Post
24 / M / Wales, UK
Offline
Posted 7/8/17 , edited 7/9/17
Predictive text doesn't become artificially intelligent just by feeding it enough data, if it did wouldn't Google rule the world by now?


Holofernes wrote:



Haven't yet heard an explanation of AI that gets past the Chinese Room problem.

http://www.iep.utm.edu/chineser/


The Chinese Room thought experiment isn't even relevant to general AI, it's specifically about whether AI would be conscious - by hypothesis, the Chinese Room is a Turing Test-passing artificial general intelligence, the only question is whether it', and implicitly also any other AI we might build, is a p-zombie or not.
6080 cr points
Send Message: Send PM GB Post
32 / M
Offline
Posted 7/8/17 , edited 7/9/17

amejia0 wrote:

AHHHHH another Fredthread :D


I eat poo.


hahaha


Fred isn't good at Chemistry
First  Prev  1  2  3  Next  Last
You must be logged in to post.