A.I. part one.

 What type of A.I. would you build?
9968a8bff23cc065dafe2452ec975ebc
In this piece I will go over a few types of A.I. and what ones to cobble together in theory.
After some realize they do not want smarter than human general intelligence A.I./S.L/A.G.I. then the real question to ask is what type of A.I. do you or others want. Most will want A.I. that serves them and by that definition they do not want it smarter than them or everyone. They really want things like search algorithms and sudo human chat bots and recognition systems and such.
It seems like most will want some form of chatbot that has deep learning of context. A chatbot with Watson like A.I.. If you hear them going for religion or emotion first then realize they have further to go and problematic in ways. I just tend to laugh at them or ignore them, as they have further to go. Some might even want personalized shopping or such. Though if you where to ask me that’s more of a harassment and broken as the algorithms have shown. For me I want A.I. to filter out ads and site suggestions as they have always been wrong similar to the astrologist or card readers and such. If they could only get to 50% it might not be so bad or annoying. Some will be after quality of content and context over volume of data. While others may want social though that seems to be a joke in and of it’s self.
To cobble together your own version’s you may want to look at some of these examples.
Watson
to a chat bot with the voice you prefer or a siri/cortiana type voice. Then link into a game engine for further construction like unreal engine. Most people will want some type of human like avatar for the A.I.
Basics of AI stream and delve into more advanced AI
You may want it to be able to call up search engines or plug into them. IE google, bing and such. That or make your own web crawler and avoid them tracking the A.I.. you will also probably want it to run in a encrypted layer to help minimize theft from spy agencies and companies like google and micro soft.
If you want it really secure then you want to build against logic sniffers and deep packet inspection when on the web. ( Not a easy thing to do but possible for some )
Some may want there A.I. to draw art or play music and customized to there preferences. ( not other companies preferences as they tend to get that wrong most the time. Examples
and
or maybe you want your A.I. to tell you poems it makes or mines from the internet and links to your preferances
or maybe you want it to make stories customized for you
or maybe even make you a game customized to you and your reactions as you play
Or play a game while you watch

digital data security

Maybe your not interested in the music art and games part and what a A.I. that does stock trading for you better than any human can do. It makes the stock market look broken in a few ways and only for a type.
or maybe you want it to mine legal information for you
Or maybe you want it to help in science or mine text’s and read them to you threw a voice to text like app
You will probably want a face recognition system like deep face
or images in general
or intrusion detection
Oh and you will want speech recognition for it to hear you with. Though like other A.I. algorithms I must stress some work better than others. Siri,  cortiana and dragon show that to degrees.
I would stress to a point, you will need to pick the A.G.I. sub parts you want to cobble together.
As you see with some of the above A.I. subsystems/systems you just need to figure out what you want in the system you want. You could even go for cyber warfare suit controlled by A.I. or drone and plane flying. Part of the push for A.I. that can play go is warfare systems. Chess was just a small step in that and warfare logic. Sun Tzu and go the next level.
 cortanaedi
 drone-pilot
Last for this post. I would stress some of these systems are not the same and you will need to work on cobbling them together and making the right weighted systems or priority calls or such. From visual recognition to speech recognition to data mining to navigation. You will have to find the systems or algorithms you like and work on cobbling them together in the right way. Some people will be better at this than others, and I hope a range of hybrid systems shows up. Even some that use some quantum computer functions for specific problem solving or recognition systems. Some things are faster or better with classical computers while others might produce better results on quantum processors. Like wise some things can be subsystems or preprocessing and no need to over task the main A.I. system. ( like human reflexes in a way ) Also if you make one in a game engine you will need to learn a few things in animation to rig the A.I. and speech or walking/running and make it look right.
rigging1
Advertisements

About laborious

Interests in science, art, math, theory, game engines, future technology.
This entry was posted in Uncategorized. Bookmark the permalink.

One Response to A.I. part one.

  1. laborious says:

    Google DeepMind co-founder Shane Legg – Machine Super Intelligence

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s