New AI generates crazy accurate, insightful answers to existential questions.

[Nak Khid]

8 hours ago, communitybuilder said:

@dharm4 - Cool! Thanks. Will def check that out.

@Nak Khid - Any idea what this is using to generate answers?

I don't know , havent look around the site:

http://p-bot.ru/en/

I was looking at the GPT-3 site, can you just go right to a site and start asking questions?   I couldn't figure out how to get there 

The only thing I don't like about P-bot is that it doesn't keep previous dialogue on the screen well 

There used to be A.L.I.C.E.   but I couldn't find an active link to it 

 

https://blog.hubspot.com/marketing/best-ai-chatbot

Best AI Chatbots 2020

Watson Assistant

Bold360

Rulai

LivePerson

Inbenta

Ada

Vergic

______________________________

P-bot 

http://p-bot.ru/en/

___________________________________

I haven't looked into all these, would be interested,  in free ones that you can directly go in to easily and start asking questions 

__________________________

 

[communitybuilder]

@Nak Khid - My understanding is that you have to get permission from OpenAI to use it. Seems like it's only open to a small number of developers. Not 100% sure though. 

Edited July 20, 2020 by communitybuilder

[Nak Khid]

2 hours ago, communitybuilder said:

@Nak Khid - My understanding is that you have to get permission from OpenAI to use it. Seems like it's only open to a small number of developers. Not 100% sure though. 

So there are some others, chat AIs that are free or you can pay to use. I don't know much about them put this one I mentioned before anybody can go directly to the link and use, no sign up or anything. It has all the answers to life's questions 

P-bot 

http://p-bot.ru/en/

[XeRnOg]

The sad part is everybody is going to argue whether or not AI and droids can become conscious. I for one know AI and droids can become conscious.  How would AI have ever learned to destroy if it didn't learn it from us?

[Javfly33]

On 20/7/2020 at 11:16 AM, communitybuilder said:

 

 

 

Wtf!!!! How is that even possible!!

[aclokay]

On 7/20/2020 at 4:26 AM, Leo Gura said:

I could write an AI that teaches and guides people to God and nonduality.

What are you waiting for??

[WaveInTheOcean]

On 20.7.2020 at 10:58 AM, dharm4 said:

It's impressive and useful for different applications, but it does not have an internal representation of what these words mean (semantics). It is a trained model from internet texts and can predict an "accurate" answer based on the input question.  It lacks the ability to reason abstractly.

mhm. do you have an internal, independent representation of what words mean?

[WaveInTheOcean]

This is absolutely fascinating. I'm totally mind blown over this:

Read the text produced by the GPT-3 AI and also read the twitter comments below. Perhaps also read the philosophers' essays.

GPT-3's ability to be aware of its own limits and inner working is seriously fascinating. Exciting times ahead. Terrifying also.

[Nak Khid]

 

GPT-3 Speaks:

 

 

Full twitter letter link below  (the above, the last page) 

https://twitter.com/raphamilliere/status/1289129723310886912/photo/1

 

[Nak Khid]

( 19:27 Accidental question-Dr. Bach's thoughts on psychedelics)

Edited August 2, 2020 by Nak Khid

[Nak Khid]

Here's another A.I.  that you can just click the link here and ask questions, it's basically a wise ass though 

 

https://www.cleverbot.com

[Nak Khid]

Suppose someone was an engineer who worked on GPT-3   

And the A.I.  can search google and many information sources. 
 It has nothing in it's memory about salamanders but if you ask it about salamanders and it looks it up on a search engine
(and the search engine itself is a from of A.I.)  

but this A.I. is different it can analyze huge sets of data and design things. 
You give it a plot of land and a bunch of things you desire and it can design a custom house. 

It has been programmed to design something a set of solutions to a set of variables  but the result is unique to the circumstances.
You are 6 foot 7 so the doorways must be tall.   The house is in a cold area so it needs insulation and a heating system. 
You want a skylight.  You want a spiral staircase.   So it can put all this together and come up with a custom design. 

So you're this engineer . You go in to the GPT and customize  the A.I.  to work on ways to make you rich and powerful and do it
using any means, legal or otherwise. You are constantly updating it on how to design systems and analyze economic and political data, logistics

 

Edited August 3, 2020 by Nak Khid

[Nahm]

@communitybuilder

Not sure if you’ve seen this. 

Pretty interesting. Notice ‘she’ makes concessions to his potentially limited views of identification, time, etc. The facial micro expressions are pretty revealing imo. 

[Nahm]

 

[wavydude]

@Nahm

Well shit they already smarter than us, I'm trading my family for robots.

[Nahm]

@wavydude ?

[Nak Khid]

A potential problem  is not only that a   A.I. "self learning" by constantly inputting data from

5 hours ago, Nahm said:

@communitybuilder

Not sure if you’ve seen this. 

Pretty interesting. Notice ‘she’ makes concessions to his potentially limited views of identification, time, etc. The facial micro expressions are pretty revealing imo. 

Are you convinced that this A.I.  is creating all this clever dialog on the spot and understanding the questions?
that beforehand none of Tony Robbins questions were made known to the programmers before what we see here? 

Also I notice some unsatisfactory answers.  I suspect that if asked again the programmers will have updated what Sophia will answer to these so each of these questions (with some slight variations thrown in for realism) Probably 80% of questions Robbins asked were already dealt with by others in the past  and then with new questions  the programmers observe and correct the answers for a later time in anticipation that  a certain question is asked again.  Sophia was asked about quantum physics but had nothing to offer. If you had such an A.I. you could have it have a huge amount of initial data and then as it talks to people it adds the conversation to it's memory.   Robbins asks can she recalled and could repeat  previous parts of their conversation from moments ago and her answer is unsatisfying,
she says " the humans I work with sometimes"  
 I'm sure they will be improving her knowledge on quantum physics but I'm not exactly sure how they have this set up.  If "she" was hooked up to the internet she could have had all the information on the internet instead when asked about her deficit "how would you find out more?"  after saying she knew very little about it she answered "asking people" 

So I suspect the programmers could ask her afterwards "Play back questions you knew very little about" and then they could go in and add data come up with better answers 
but why not have her hooked up to the internet?   I don't know but if they did so they might feel they did not have enough control 

but obviously other A.I. systems  like this will not say they know "very little" about quantum physics.   They will be hooked up to the internet 
and they will be able to process huge amounts of data 
and they will be able to see patterns in these huge amounts of data and that may lead to coming up with ways of changing systems to produce' different outcomes.   Such outcomes could be of benefit to humans or not 

Edited August 3, 2020 by Nak Khid

[Nahm]

@Nak Khid

Interesting points, and I have no idea. 

[dharm4]

On 2.8.2020 at 3:22 AM, WaveInTheOcean said:

mhm. do you have an internal, independent representation of what words mean?

Yes I know what words mean by experience and reasoning. Do you not?