We welcome Ben Stern (Manager of Educational Partnerships at TeachBoost) as our guest on the Mouse Blog. His comments below are his own, i.e., they do not necessarily reflect the opinions of either TeachBoost or Mouse; TeachBoost and Mouse have no relationship.

Daniel: Ben, in your Sept. 11th EdSurge post "Software Will Not Eat Education" you assert: "...many issues in K-12 [...] are not software problems. They are human issues, political issues."

If so, how do educators, entrepreneurs and investors differentiate which issues are most appropriately addressed with software and which with other means? And, to the extent that software is a human construct, how do we truly separate the dancer from the dance?

Ben: Daniel, first of all, thanks for following up on the article with these great questions. It’s always a pleasure to speak with you and Mouse.

Great question here. Let me take a stab.

All important problems are human problems. A computer can’t tell me how I should be living my life. Similarly, if I’m reading To Kill a Mockingbird with a group of students, a computer cannot effectively lead a discussion that draws on students’ interests, personalities, and our relationship cultivated over a whole year to get every student to grapple with central themes in their own way. Software can help me better understand their interests, or allow me to surface trends across their work, or allow me to connect with them outside of my 45-minute class period. But it cannot replace me, the teacher, in the most important part of the whole teaching process. Good teaching relies on non-rational, non-computational, distinctly human elements like relationships, personality, and instincts that cannot possibly be digitized. How to teach well is an important, difficult, yet partially unresolved problem.

I argued that political problems are also beyond the scope of software. The access to education provided by technology may play a role in closing the achievement gap, but learning is more than listening to lectures and typing up answers to questions. MOOCs will continue to have no real impact on K12 education. Software on its own, without legislative and infrastructural changes, isn’t about to resolve the disparate quality of education among all American schools entirely. Education reform will happen through policy and legislation first and foremost -- would you trust a law written by IBM’s Watson? Would you trust Watson to teach your kids?

Daniel: You also state that companies should focus on narrow sets of problems and recognize that they operate under constraints. Isn't that a bit of a straw man, i.e., don't ed tech (indeed, most) companies understand the need to focus, focus, focus and know quite well the confines against which they collide?

Ben: Well, hopefully! My thinking here was that I see a lot of software that overreaches, that fails to address most teachers where they are today. There are some amazing, innovative teachers who speak at conferences and have huge Twitter followings. But these teachers are outliers. Just like we see a lot of startups coming out of Silicon Valley whose product only gains traction with the technology elite but never with the general public (and die accordingly), we see plenty of edtech built for the Twitter community, not for the more typical American classroom. Yet, the latter is exactly where edtech is needed and can help the most.

Daniel: Aren't you being overly cautious in suggesting that ed tech companies seek iterative, not disruptive, change? After all, those who in the late 19th and early 20th centuries created the foundations of today's American K-12 structure were hardly shy about proclaiming a revolution. For instance, Dewey said that "every great advance in science has issued from a new audacity of imagination."

Ben: Sure, but did Dewey disrupt education? Reading him today is a lot like reading Edutopia or BIE’s publications. It’s great stuff, and I agree with it. But the fundamentals of teaching and schools, the actual act and its relevant institutions (as opposed to the content), really haven’t changed all that much for thousands of years, much less these past two centuries. They’ve evolved, but they haven’t been disrupted. I’ll double down on my claim from my article that iterative, software-driven change in education is ambitious.

Perhaps in dispute is what we mean by change. When I hear Marc Andreesson say software will eat education, I believe he means a fundamental rethinking of schools and teaching, with software becoming essential to and even replacing both. He’s talking about the end of school buildings, of class periods, of 20:1 teacher-to-student ratios, of grades, etc. That would be disruption. Besides the clear obstacles to such fundamental change - there’s a lot of money, legislation, infrastructure, institutions, expectations, other industries, etc. tied up in this current model - it’s not evidently good to replace it. Education is a different industry than, for example, transportation. Disrupting the taxi industry makes my life considerably better; Uber is disruptive and for the good. Would a software-based, teacherless K-12 education improve our lives?

The institution of a school and its practices has worked in varying degrees for many centuries. Of course, the content of what students learn has, will, and should change, but that sort of change isn’t really the sort of disruption that Andreesson or we are talking about. Rather, the edifice of education as we know it, everything besides the content of what we teach, will probably not change completely. In my view, it shouldn’t. We can use software to improve it, to iron out inefficiencies, but not to replace it. I’m advocating for iterative change, like we already have seen over the centuries in education.

I’ve received some really strong responses to my article, particularly on this point. While most have been positive, some argue either that I’m not dreaming big enough, or that the irresolvable nature of these problems by software indicates that the whole model needs be blown up, to be disrupted.

To the former point, I’ve just responded that I never said innovation is impossible in education. It’s possible -- it’s already happening. I deeply believe that the company I work for, TeachBoost, is making a meaningful impact on schools using software. Streamlining educator effectiveness programs and taking something painful and unproductive (teacher evaluations) and turning them into opportunities for professional growth is innovative. But we’re not replacing instructional leaders; we’re empowering them. That’s innovation, not disruption.

Innovation is different from disruption, and disruption is what I wanted to critique in my article. The dogma of disruption - faith in the inevitability and inherent good of disruption in any and all industries - is a bit dangerous. 90% of startups fail for a reason. And the VC bubble that’s formed around this dogma makes it harder for truly innovative - but not disruptive - companies to compete. I do think software will continue to fundamentally change a great deal about our lives in good ways, and for the most part believe the rise in entrepreneurship and innovation generally is a good thing. But we need to maintain reasoned, cautious optimism. “Disruption” and “software eating the world” make for great rhetoric, just as Dewey’s call for a “revolution” continues to inspire. But those of us in the trenches fighting for our companies need to be mindful of certain realities that our rhetoricians don’t.

Daniel: Thank you for your insights, Ben.

Daniel Rabuzzi is Executive Director at Mouse .