Thursday, 26 February 2015

Future of Ed #AI DeepMind computer teaches itself to play Atari 2600 games

How do we see the world when Artificial Intelligence takes over work that we find 'humanly satisfying'? What about game-based learning? Is it really such a game-changer, and if it is, than what do we think about computers teaching themselves to be better (than us humans in these games)? Education might get under pressure, as automation and learning creeps in from the digital world. Will Art be the last job standing? All of these ideas are fine with me, as long as we readjust society to cope with these new, upcoming - and more than often based on learning - realities. Can we rethink our place as humans in this increased technological world, and in such a way that we will all still be able to have a sense of intellectual satisfaction? Or will we get globally depressed once we acknowledge that our brain can be outrun by any improved AI DeepMind next generation computer or algorithm?

DeepMind takes over Atari high scores 
Google's DeepMind is something to be reckoned with, and an technological evolution that will push  us to rethink society (at least, that's what I think). The DeepMind algorithms are very interesting, as they can be seen as self-adjusted learning/teaching algorithms. DeepMind is used in DeepFace in the form of DeepLearning (earlier blogpost on it here), and now DeepMind cracked the teaching/learning code to get better at simple video games (yes, this does relate to the tech teenage movie WarGames, the movie from 1983) as The New Scientist (and others) reported today.

Announcing the next evolutionary step: a tiny baby now, but growing
This is a great breakthrough in technology, but potentially one that can influence learning/teaching and society. First of all, I am all for it. Evolution is needed, especially when looking at the boundaries of humanity (evidence-based research seldom results in other more durable approaches (think climate, hunger relief...), war is still a major driving force although we all know the downsides of it). So in a way, my hope rests in the next phase of existence, which might just be fully digital, with us humans as reservation or zoo kept animals, that are provided toys just like our closest biological sisters and brothers, the apes.

Education as savior to get us from human-biological to computational-digital  
But there is a downside as we are in a transition zone, where AI is not yet capable of taking over, and humans are decreasingly needed. The shift from the biological to the computational needs to be made more pleasant. Allowing people all over the world to be have their little piece of Eden, before the fully computational revolution takes over. Education can help, especially lifelong learning, as this will allow us to re-evaluate which opportunities are still open to us humans, and how (or where) we need to turn to reshape our knowledge, or take on a new identity that will allow us to live a satisfactory life in this transition phase between the human-biological and the computational-digital life. Education would as such no longer need to be job-focused, but more life-focused: sustaining and supporting that which makes us humans feel satisfied, intellectually balanced, good.

Online learning as one of the tools to guide and shape us
Elearning or online learning with multiple devices, across location and time is a good option to re-adjust life to fit these new, upcoming artificial intelligence changes. At best, online content is shaped by cooperatively working on a particular subject. Multimedia files, content, contextualized authentic learning experiences... all of these can be brought together much quicker then ever before, and built by all of us putting our heads together (standing on the shoulders of giants comes to mind). So, in a way, online learning can offer quick responses to new societal changes. Providing new opportunities, new ways of looking at the rapid changes, and at our human identities against the backdrop of a changed, more AI oriented society.

(image source: