Aiko BRAINS software a closer look.

Hi

This is a screen capture of the BRAINS software. Hopefully, it will make more sense after watching the video.

Wow that was great and so cool, I learned a lot from that video.

Could you and would you ever create a subroutine to have Aiko ask you questions?
Like why, how, what is that, what are you doing and so on. Could you then put what she learns into the database?

I’m glad to see that you concord the font size problem. Does that problem have to do with zooming and focusing?

Thank You Le for that video.

Kthomp06 wrote:Wow that was great and so cool, I learned a lot from that video.

Could you and would you ever create a subroutine to have Aiko ask you questions?
Like why, how, what is that, what are you doing and so on. Could you then put what she learns into the database?

I’m glad to see that you concord the font size problem. Does that problem have to do with zooming and focusing?

Thank You Le for that video.

That would be kind of hard to do as it takes several factors that must be already programmed in her.
Akio is still pretty much a baby as far as education goes.

If I tell a child the sky is blue, he'll ask why, but until he knows the sky is blue, he won't ask why.
I like the fact Le put her in two accents as well as languages.

Artilects rule! wrote:

Kthomp06 wrote:Wow that was great and so cool, I learned a lot from that video.

Could you and would you ever create a subroutine to have Aiko ask you questions?
Like why, how, what is that, what are you doing and so on. Could you then put what she learns into the database?

I’m glad to see that you concord the font size problem. Does that problem have to do with zooming and focusing?

Thank You Le for that video.

That would be kind of hard to do as it takes several factors that must be already programmed in her.
Akio is still pretty much a baby as far as education goes.

If I tell a child the sky is blue, he'll ask why, but until he knows the sky is blue, he won't ask why.
I like the fact Le put her in two accents as well as languages.

I don’t know if it would be so hard, Aiko already recognizes images and faces. I think people make thing harder then they really need to be.

Example: I’m assuming here that Aiko has a picture index. Could Aiko look around the room and find a picture that is not in her index? Could she not display the picture in the picture box? Could she then select a case statement to ask the question “what is that” or “What is this item”? Then upon receiving the answer, store the picture to the picture index and the answer to the database with corresponding picture index key. If she doesn’t receive an answer she could file the picture as unknown for later inquiry.

So to me Aiko is already doing this. When Le asked Aiko “who am I” she took a still picture then some word appeared in the blue screen picture box as she scanned her index, and then she gave her answer.

So I’m curious, what are these several factors you mention? Maybe I have missed something.

If you tell the child the “sky is blue” after he/she has asked you “what color is the sky”. How do you answer if the child asks “is the sky always blue”?

I'm just looking to see if it could be done and how one would go about doing it.
More input, need more input. (Johnny 5)

In 20 years, BRAINS and Yumecom will be household names

Hi

To: Kthomp06

What you are saying make sense.
That is how the Brains software process people and objects.
"Through that square box".
The problem is the process of learning new object.
In theory, Aiko can scan around the room, any object that she doesn't know she can learn it, but
each object require about 15meg of data.... so it will add up pretty fast.

Kthomp06 wrote:

Artilects rule! wrote:

Kthomp06 wrote:Wow that was great and so cool, I learned a lot from that video.

Could you and would you ever create a subroutine to have Aiko ask you questions?
Like why, how, what is that, what are you doing and so on. Could you then put what she learns into the database?

I’m glad to see that you concord the font size problem. Does that problem have to do with zooming and focusing?

Thank You Le for that video.

That would be kind of hard to do as it takes several factors that must be already programmed in her.
Akio is still pretty much a baby as far as education goes.

If I tell a child the sky is blue, he'll ask why, but until he knows the sky is blue, he won't ask why.
I like the fact Le put her in two accents as well as languages.

I don’t know if it would be so hard, Aiko already recognizes images and faces. I think people make thing harder then they really need to be.

Example: I’m assuming here that Aiko has a picture index. Could Aiko look around the room and find a picture that is not in her index? Could she not display the picture in the picture box? Could she then select a case statement to ask the question “what is that” or “What is this item”? Then upon receiving the answer, store the picture to the picture index and the answer to the database with corresponding picture index key. If she doesn’t receive an answer she could file the picture as unknown for later inquiry.

So to me Aiko is already doing this. When Le asked Aiko “who am I” she took a still picture then some word appeared in the blue screen picture box as she scanned her index, and then she gave her answer.

So I’m curious, what are these several factors you mention? Maybe I have missed something.

If you tell the child the “sky is blue” after he/she has asked you “what color is the sky”. How do you answer if the child asks “is the sky always blue”?

I'm just looking to see if it could be done and how one would go about doing it.
More input, need more input. (Johnny 5)

It would be amazing to figure out how the software could make context into a question,
Better yet turning information into a story.
How would one know what a sky is,
Or when one is in a room , box or cave.
What is the key ingredient to define the space around us.
How do we perceive it should vision be enough is the ear balancing the sheer force of the earth making me aware on a subconsious level that there is a space above , under in front and back.
More over how would it benefit the knowlegde about anything this is typical Godel, Escher Bach.

Would it be that Aiko has sensors for picking feedback of the world more like accelerators and gyro's or software functions measuring light and other enviroment processes?. I'm very amazed by the data consumptions Aiko has.
All that without compression gives the impression she consumes more then data but also computing time.
Does Aiko handles multiple processes to keep track of anything simultaneous.

To: "WulfCry"

Very interesting topic you just post up.
How do people know when they are inside a room or outside or in a cave for that matter.
When you are inside I guess you look for wall, but when you are in a cave.. is that outside or inside.

To program something like that is very tricky, and I don't know how to do it.

The vision is uncompress, that is why the file is large.
In the future, maybe I will make it compress file.
But making compress file will take longer to process and require more cpu power.

Mutitasking,
Yes, the software does do multitask, but not the vision.
Task 1 - for vision
Task 2- speech, recogniton etc
Task 3 - hardware.
Task 4) unused but the resource should be left for OS.

I set one threads for each task.
The max core2 is 4 threads.
However, the new core7i has 8 threads.... so idealy, it should be faster, if you split the task more.
Personally, I think the vision should have at least 3 tasks.
1) extract the background from foreground....(always keep constantly doing every seconds)
2) extract the important object and compare to background
3) process images into database and comparing it.

This way there is no frame freeze, as you seen from the video... alot of frame freeze..

Within just five years, what you said will be SO out of date that you'll be all like "I said that? ME?!"

While it may be difficult to impossible now, with the rapidly upgrading tech (I heard somewhere that tech has progressed more in the last decade than all of the previous 5000 years combined) it could be a simple solution within just a few short years.