Mothering Forum banner

1 - 6 of 6 Posts

·
Registered
Joined
·
3 Posts
Discussion Starter · #1 ·
Hi.

I research artificial intelligence and have great knowledge, but would like to ask new moms (or moms) a important question:

Once your baby learns to crawl, it should reach a wall in the house. And I'm talking about the first time ever it reaches a wall by the method of crawling. Does your baby continue to crawl into the wall? How long? Or does it (on the first time) reach the wall (see or touch) and turn without learning to turn? If it continues to crawl into the wall (for the specified amount of time you give) then the baby should stop at some point, cry, and start randomly moving.
 

·
Registered
Joined
·
3,284 Posts
Wow. You've never been around children, have you? They're not a windup toy. You don't set them down and they go in a straight line until you move them. Babies move toward things that interest them. Long before they do a proficient crawl, they lunge, scoot, wiggle, roll toward the object of their interest. Their first crawl is more of a rocking, falling lunge toward what they want. Some babies never crawl. They butt scoot or pull to stand and furniture cruise. A normally developing baby would never crawl into a wall and get stuck.
 

·
Registered
Joined
·
3 Posts
Discussion Starter · #3 ·
I know when they crawl they swerve.

I know they probably learn other moves that get them acceleration first.

But the question still stands - when they roll or etc to a wall, do they keep trying the same motor-direction-actions into the wall and then start to cry and randomly try new random actions? Or instantly say roll the other way away from the wall?

My question's point is that they learn actions to move but if they get stuck at a wall they will either instantly turn away from it/do the opposite actions OR start trying to come up with new actions.

Does anyone here know which one happens?
 

·
Registered
Joined
·
3,284 Posts
I've never seen a child run into a wall like that. (When they're older they might if they're not looking where they're going, but that's not what you're talking about).

Again. Babies go to what they want. A plain wall does not draw their interest. You're not going to find the answers you seek. Humans aren't like machines.

But, do babies get stuck sometimes? Like a leg in the bars of a crib? Or their head between the banisters? Sure. They usually try to extract themselves first and then cry for help if they can't figure it out. I doubt any mom can remember the very first time their hand got caught up in the blankets and what exactly the child did.

Child development is not linear anyways. It happens in fits and spurts, erratically over time. Even if the baby remembers how they removed the hat from over their eyes last time, they might not the next time.

So, the answer to how babies learn to overcome obstacles is in a swervy erratic line with lots of failures, lots of help, trial and error, good days and bad days in an extended 20+ year dance to adulthood.
 

·
Registered
Joined
·
3 Posts
Discussion Starter · #5 ·
Well, by the sounds of your reply, you either don't remember or didn't see what I said in my last reply - when the roll/crawl gets them to a wall for the first time.

You see, the actions are generated randomly, and when the accelometer activates higher it finds better actions for all motors (being 1 set tried that is better than the other sets), then to say crawl it is many sets ~ left only left arm, then only right, hips forth, etc. Now, while the actions to crawl/roll are tweaked as used such as swerving-crawl, they still just do the actions on cue, so they should reach a wall, and either will do pre-installed turn actions when see wall or start generating actions once at the wall. Though if they never have had enough room (crib) or no walls nearby they very well could once a few months older jump the bridge and like even be told what to do at the wall.

Kinda need someone to (unless remember) to really watch n see what happens....any new moms/soon to bes here looool? You could contact me later!

All that crying/laughing is automatic pre-installed actions done when get pre-installed +/- rewards such as a specific bad human face or good food. While the rest of the system is, well, automatic, in a more complex way. Of course they will laugh and cry from any idea later, simply the rewarded food linked to the video game vision of what the win will be makes it a reward for actions, and it improves.
 

·
Registered
Joined
·
1,705 Posts
OK. Obviously the OP isn't experienced with kids. I'm reviving this one because it's just so darned intriguing to me.

OP, you were asking the wrong questions. They can't be answered because they contain inaccurate assumptions about how children develop. The initial mode of movement varies from child to child. Lets call it the primitive mode because I don't know the correct scientific terms off the top of my head. Crawling isn't the primitive mode. Primitive mode has the child laying down. They may scoot forward, backwards, roll in both or only one direction, or some other kind of shifting movement. My most recent baby could roll only one direction at first.

Second, the child will only be motivated to move if something interests them. So, babies rarely interact with walls because they typically aren't very interesting. Many children will not interact with a wall until well after they begin walking. Even then, they don't run into walls unless they aren't looking where they are going.

I do understand what you were hoping to understand though. My youngest found our couch interesting. So, he would roll to it from an early age. Because he could only roll one direction, he would continue rolling into it. Eventually he would become frustrated and cry.

He wasn't trying to get past the couch though. If he wanted past something, he went around it. If there wasn't a way around it, he wasn't aware that there was anything past it.

Human children will not just move in a random direction until they discover an obstacle. If you are working in AI, you need to understand that programmed robots will behave this way..... But a learning intelligence has a reason for EVERYTHING it does. If a child will be aware of a physical obstacle before coming into contact with it. They may stop to think, or take immediate action. Their action will either be to give up and do something else....Or to device a way to go under, over, around, or through the barrier. (For example, the same child can now climb over baby gates, or get a running start and crash into them so hard they break. He just turned 2)
 
1 - 6 of 6 Posts
Top