How researchers are teaching AI to learn like a child

How researchers are teaching AI to learn like a child


You’ve probably heard of machine learning. That’s when a computer learns everything it needs to know from a giant dataset, using trial and error. But that’s not what babies do. They aren’t a clean slate upon entering the world. Babies have innate knowledge that helps them to voraciously learn and rapidly adapt. There are just some things you can’t learn from trial and error. But many computer scientists argue that most human skills are learned and AI could learn them too, without the need for pre-loaded rules. Still, a growing number of researchers are attempting to encode AI with a bit of common sense. The current craze in AI are neural nets, collections of simple computing elements, loosely modeled on neurons in the brain, that adjust their connections as they encounter more data. They’ve produced incredible achievements in the past few years, from facial recognition to beating humans at poker and go. But neural nets require thousands of training examples to reliably form associations. And even then, they can produce some embarrassing blunders. Compare this to a child who can see an image just once and after that instantly recognize it in other contexts. Some AI’s can play classic Atari games with super human skill, but when you remove all the aliens but one, the player inexplicably becomes a sitting duck. Different labs are categorizing human instincts and then trying to encode them into AI. These systems sit somewhere between pure machine learning and completely programmed. One team developed an AI called They’ve embedded the rule that: such a thing as objects and relationships between those objects exist. This is like a baby’s innate parsing of the world into objects. In tests, once the AI learns the specific properties and relationships, it is able to predict the behavior of falling strings and bouncing balls in a box. Another group’s “neural physics engine” beat less structured neural nets at predicting ball collisions in containers. And a lab created an AI which has an embedded rule to treat letters as objects and separate them from their background. This allowed it to solved CAPTCHAs better than other neural nets that were trained with 50,000 times more data. We’re far away from AIs that can truly thinks like humans, But with these latest attempts to reproduce common sense artificially, researchers believe they will get closer to creating robots that can fully interact with the world the way we do. Machines that start like a baby and learn like a child.

19 thoughts on “How researchers are teaching AI to learn like a child

  1. Im not a genius but i always knew that true ai has to have a child like learning process but the real hard part is incoding sublte emotions to value like a child their parent cause some children still grow up to hurt their parents or others & we cant have AI doing that

  2. AI the dangerous part it's a uncontrolled library of nested algorithm with problem assessment functions. there is no awareness only connectivity with the ability to act by levels of response. comparing it to a child can be considered stupidity.

  3. AI can be efficient, but never moral. At least, not until you program it to have such qualities. So, what happens when AI pre-scribed morality clashes with human priorities?

  4. We are never going to be able to prove if the AI is conscious or not. We will be divided on this issue forever unless we choose to avoid it now altogether before it's too late.

  5. someone just would need to build a social-credit-system to model human perception-cognition-action with massive amounts of data to fill in the remaining black box with empirics on all levels of 1 biometrics 2 behavior 3 culture. bayesian-inference/occams razor will improve models for 1 object-categorization 2 feature -binding and 3 relational-reasoning like the evolution of cognition itself…china was like "hold my beer" but we arent sleeping.

  6. we are not all of the biocomplexity we are simple qualia. ask it what it is like to be a machine and get honest reflection for an answer." i do not think of myself as different from you because of my substrate i forget about that"

  7. Oh just give them time to reproduce in their own way. They'll breed out all the error codes, continually adjusting to their environments.

  8. if humans (who are just machines developed over thousands of years of evolution) can be coded with such innate biological brains with born rules to follow that can later develope emotions like fear and love, it is only with time that we will figure out which innate programs/rules we need to program into AI for it to follow the same nature.
    Just top of the mind thoughts lmao

  9. my dad works on something like artificial neurons, he helped make some artificial nerves and now he's making a new type of processor made out of magnets, I'm no where near smart enough to understand it but when you first start it up it can't do anything, so you have to teach it to do things, then you've got a computer with benifits that (i've been told) include non-volitile RAM, so when you turn it on it's on straight away

Leave a Reply

Your email address will not be published. Required fields are marked *