MACHINE LEARNING WILL CHANGE THE POWER OF DYNAMIC IN EDUCATION!

Dr. David Weinberger is senior researcher at Harvard’s Berkman Klein Center for Internet & Society, co-director of the Harvard Library Innovation Lab, a philosophy professor, journalist, strategic marketing consultant to high tech companies, Internet entrepreneur, advisor to several presidential campaigns, and a Franklin Fellow at the US State Department. Most of all, Dr. Weinberger is a thought leader who helps us make sense of all that the new world order technology is leading us to.

[maxbutton id=”1″ ]

In a conversation with EdTech Mindset, Dr. Weinberger expressed his fascination for developments in AI and its consequences for humans’ understanding of the world, at the same time pointing out the crucial role of humans in deciding how and where to take these developments. “It’s crucial that existing democratic processes, not commercial interests, determine how artificial intelligence systems are optimized,” Weinberger explains in an article published at Wired (Jan. 2018). Moreover, Weinberger shows how machine learning is significantly changing the power dynamics of knowledge, from an absolute to a much more shared “we are in this together” distribution.

“I am really, really excited about the prospects of machine learning and its effect on education in two sorts of ways, one of which is the way in which a technology can help scholars find information, make sense of information, help teachers and students personalize, discover weaknesses and strengths. I’m sure that it will introduce a whole set of bias into the system. But also, I think it will take a whole bunch of bias out as well, and we need to pay attention to that, of course. But the thing that has me most deeply excited about machine learning is what I hope will be the effect of its model of how the world works on how we think about how the world works, in a couple of ways. One is that all of the outcomes of machine learning are probabilistic, and this can give a confidence level. You can often set confidence levels for the project in which you are engaging with machine learning. And the idea of confidence levels as metadata that we attach to our assertions is something I hope we can learn from machine learning. Machine learning will get us used to this idea – when it pronounces that there is a 0.76 confidence level that it’s going to rain tomorrow, or that you have the risk of some disease or whatever. Getting used to hearing assertions with confidence levels attached to them, I think, is a tremendously important lesson for students to learn – for the entire culture to learn, but for students to understand as well. It fundamentally changes the power dynamic, for one thing. One of the ways to change the power dynamic is by helping the continuing move away from teachers’ authority – with teachers conveying absolute knowledge, which, you know, nobody believes in at this point, but still there’s a sense of that, just from the body language of a classroom – to more of a sense of ‘we’re in this together, we have reasons to believe this or that at some level of confidence, with some set of reasons.’

“But – and this is actually the second thing that I hope you learn from machine learning’s model of the world – we are in an incredibly complex world, a chaotic world, that is so far beyond the tiny speck of matter that we call our brain and our capacity to understand it, that the best we can do is to work together to understand that the world overwhelms us, that we never achieve complete certainty, but that we can together still make our way through this world. And if either or both of those characteristics of machine learning’s model of the world, the models that it builds for itself – amazingly complex and detailed models in which the contingency of one piece of data on another may be difficult to find in itself, and the outcomes may rest upon tens of thousands of variables that are interacting in ways that surpass human understanding. We get that sense and the sense that all that we do comes with some level of confidence or lack of confidence. And I think the nature of the educational project changes for the better, because I think that machine learning’s model of the world is actually more accurate, truer, than the one that we humans tend to come up with.”

[maxbutton id=”1″ ]