Imagine you're thinking of putting up some shelves.
The first problem is how on earth a system would know that you are planning to do that. Would it figure it out from your online shopping activity? Would it look at your diary? Would it have your mobile phone mic permanently switched on so it could monitor your conversations?
Assume there is a way of doing this. How would it know which YouTube video you would rather watch? The one with the bubbly hapless presenter or the more mature gentleman with an impressive tool-belt?
Assume there's a way of doing this. When does it send you the learning content? When you are driving? When you are in a meeting? When you are watching a movie?
Assume you can solve all these problems. There you are, sitting on the sofa, twiddling your thumbs and up pops a notification 'It looks like you're thinking about putting up shelves. Here's a video we've specially selected for you.' Awesome, right?
No. It's just bloody annoying.
In 99.9% of cases you're better off with a really good learning UI: 'find what you want, when you need it'. I think the only semi-feasible use-case for adaptive learning I've come across is contact centres: a situation where your activity is, literally, controlled minute-by-minute. And there we have it; when people think about adaptive learning they usually hold an organisation-centric model of the world, one in which activity is controlled and learning prescribed. A 'top-down' view. But the world of learning isn't like that: I encounter challenges and I choose how to tackle them. If we wanted to control the course of peoples' learning we would control the challenges they are presented with, not the learning content - e.g. with an adaptive goal-setter (a bit like the health apps or, frankly, games which bump you up to the next level when you get good.)
There's actually another, deeper, reason why adaptive learning is doomed: imagine someone offered you a device that sits in your car and - based on your location - prints out a map of the surrounding area. Neat, huh? I'm guessing you'd stick with your SatNav. Adaptive learning might be cool if we didn't know how to automate - to give people guidance. Consider a system that supports traders: it sees the oil price spike and says 'here are some interesting articles about oil to read'. Ok. But wouldn't it be better to have a system that says 'now is a good time to buy oil'? Yes, I know - less learning would take place. Less capability would be built. Welcome to the world of automation, in which learning belongs to the machines now.
N Shackleton-Jones , como casi siempre no solo él sino los investigadores anglo-sajones, hablan siempre de un APRENDIZAJE ADAPTATIVO, este aprendizaje lo puede hacer cada persona pero siempre con unos objetivos comunes, o sea, es unformizado igualmente y jerarquizado, pero no lo hacen nunca del APRENDIZAJE PERSONALIZADO, es más no lo contemplan generalmente ni en sus (investigaciones), ya pasan directamente al Social learning,
ReplyDeleteEn este escenario los LATINOS les llevamos mucha ventaja, auque ellos ni se enteran, ni saben que nosotros investigamos el Personalized learning y el Social learning como una continuación inseparable uno del otro y es aqui donde deberiamos hacer valer nuestras propuestas, lo cual, como siempre, no hacemos...juandon
Hi Juan. You make an interesting point about a distinctly Anglo-Saxon focus on adaptive learning vs a Latin bias towards social learning (if i understand you correctly). It is a very mechanistic, individualistic way of solving a problem - the machine sends you the content; you are isolated. I would agree that thinking about learning in a social context is better than the content-centric ('content dumping') model I usually encounter: in a social context the affective dimensions are imminent to the learning.
DeleteHola Nick, El mismo debate lo tuve con Stefen Downes en Caracas, hace un tiempo (https://www.youtube.com/watch?v=saWaGW7sJnQ) y debatimos de lo mismo, bien al cabo de un tiempo veo que el mismo Stefen ya esta pasando a posicionamientos personalizados y más sociales. Esa es la línea de mis propuestas.
DeleteSiempre he dicho que necesitamos escenarios de confluencia Latinos-aglosajones sobre ello, ya que a veces queremos hablar sobre lo mismo, pero parece que lo hagamos de cosas que nada tienen que ver.