So, apparently a team of Georgia Tech graduate teaching assistants working IBM’s Watson platform managed to produce a glorified interactive FAQ for their 300-person online “artificial intelligence” course and now we can look forward to a future without teachers.
Pardon my snark, and no offense to the technical achievement of creating the Georgia Tech algorithm – I mean, virtual TA, Jill Watson – but I’m searching for an appropriate and effective rhetorical weapon to puncture the hyperinflated futurist bullshit that infects so much ed tech-related journalism, and seems particularly pronounced at Quartz, which championed Jill Watson’s virtues in a breathless recent article.
As “thousands of questions at a time” “piled up” Prof. Ashok Goel tasked his TA’s to develop an automated bot to reply to student requests to, for example, re-explain a concept, tell students where and when the study session is located, or when an assignment was due.
I’m wondering at what point Jill Watson started replying, “It’s on the syllabus.”
Prof. Goel himself describes Jill Watson as a “minimal question-answering” tool, and yet this doesn’t stop Quartz from positing a future where “bots could be integral to college.” “They could be your tutor, advisor, grader – essentially the bulk of teachers’ work.”
The article has a couple of nods to evidence contradicting the AI-education narrative, including the failure of existing adaptive software initiatives to do anything other than train students to use the software, and a de rigueur quote from a skeptical human educator.
But the overriding impression left in this flavor of tech writing is that it’s only a matter of time before this stuff “works.”
Never mind the more than a century-long legacy of what Audrey Watters calls “teaching machines” failing to fulfill even a smidgeon of their promises, whether or not it would be a good thing to replace humans with automation – even if the automation could do everything we hope for – is ignored almost entirely. These claims for the transformative nature of education technology are often built on un-examined ideologies.
As history has shown, teaching and learning is not a technical problem to be solved, but a social process to be engaged with. The Gallup-Purdue Index on outcomes of college graduates demonstrates experiences of mentorship and personal connection with faculty seem far more important to future economic security and personal well-being than what is studied.
Education is a process, not a product. What happens to the process when more and more genuine human interaction is substituted with bots?
Let’s consider some of the questions Jill Watson must have received, such as the ones about deadlines or class resources. Human instructors often must deal with these questions and sometimes they are deeply annoying as we wonder why students are asking questions they possess the ability and resources to answer for themselves.
When this happens to me, I politely reply to students and point them towards the source that will help them satisfy their curiosity, e.g., it’s on the syllabus. One of the roles of the instructor is to prevent students from developing a pattern of learned helplessness by asking them to take a beat and see if they can answer a question for themselves.
This is doubly important in a writing class because we are moving towards becoming self-regulating as writers, developing the necessary self-critical faculties to evaluate and improve one’s own work. If I encourage students to defer their judgment or agency to my authority, we’re never going to get there.
Believing in this sort of AI as a suitable replacement for human labor is to define down both the role of the instructor and the responsibilities of the student. This is why I strongly caution students against using “grammar check” tools – very crude AI to be sure. They need to develop control over their writing and outsourcing that work delays that process.
Besides, grammar checkers don’t work for crap anyway.
Reading these breathless articles has me thinking about a bigger question that I’m having a hard time answering:
What is the most successful education technology of all time?
I recognize that my use of “success” is vague, but perhaps that vagueness suits the discussion as we consider the different categories of “success.”
The personal computer and the internet have had a profound impact on education, but they are obviously not strictly education technology, but instead are examples of technology used in education.
Many education technologies have been widely disseminated, such as the learning management system, or the smart board. From a market penetration/sales standpoint these things are “successful,” but at least in the case of the smart board, despite its significant penetration, they’re barely used, and for many of us no more useful than a blank wall.
And has the existence of the LMS correlated with improved student learning? Could it even be having a negative effect on students?
Maybe I’m missing something because it’s not my field – I mean, I must be missing something – but I’m having a hard time naming a purely ed tech innovation that has had a significant (positive) impact on education, let alone being transformative.
The history of education technology is far more one of unfulfilled promise, though in just about every case that promise – MOOCs, cough, cough – was built on almost pure hype .
As ed tech heads toward becoming a $250 billion per anum market by 2020, I keep wondering what the hell is it we’re buying.
 As Watters reports, in 1913, Thomas Edison predicted the “motion picture” would make books in schools obsolete.
 A grammar checker would flag this sentence, but it’s one of my absolute favorites of the entire post.
 Here I’m thinking about the potential of an on-demand, consistently updating grade portal to drive student anxiety.