I missed that you responded to me. I’ll try to clarify my second point.

I am not talking about a sunk cost that teachers are loath to give up on. Instead, I am positing that many teachers would be very willing to give up the TI calculators if they thought it was in the best interest of their students but, in my experience, teachers are almost always realists by requirement of the job.

Teachers are not convinced that moving away from TI’s is in the best interest of their students because of things like the SAT. They know that students who are skilled with the calculator will have an advantage on the SAT. But they also know that using the TI efficiently requires a substantial investment of time and can only really be achieved with fairly regular use. So teachers ask students to use the calculator. Test-mode (which is AWESOME by the way) and making continued inroads with test makers will answer the empirical question. If I was in the classroom, I’d be very tempted to completely dump graphing calculators and to switch to Desmos with test-mode.

And the happy ending to the story is that, like you, I have now joined the edTech field and have created a new tool that allows blind students to perform arbitrarily complex algebraic manipulations – a much higher calling than selling buggy whips.

BTW, I see that Desmos now offers UEB and Nemeth Braille output for its calculator – very cool! Also love the slide-whistle feature.

“You can get me to do algebra. You can even get me to do calculus. But graphing is where I draw the line!” – Wish I knew who said this!

My objection to the TI is the simple fact that outside the school environment they are non-existent. Nobody that I know of that uses mathematics on the job (other than teachers) has a TI on their desk. (I know an engineer that has a slide rule on his desk but he is older than dirt. I think he has it just to freak out his younger colleagues.) It kids use nothing but the TI then they are being shorted.

As for the larger picture, yeah, it’s obviously standardized tests. Not so much the ACT/SAT, as a graphing calculator really doesn’t help that much over a scientific (I looked at the practice questions on the ACT website, and only saw one question where a graphing calculator was the easiest way to solve. Sure, there were another three or so where it was a possible way to solve, but if you knew enough to solve those with a graphing calculator, you probably knew enough to solve them without one as well.), but the AP tests, which absolutely require one. It killed me last year to enforce a graphing calculator instead of Desmos on my AP Calc class, as they were struggling with both calculus and the use of a graphing calculator at the same time. But if the goal is to get them ready to take the AP test, then they need to learn to use a graphing calculator to be ready for that test, which is not a thing that can be taught in a week (Dan, I think that was Eric’s second point, if you want them to learn to use a graphing calculator well enough for it to be useful on a test, that level of fluency with the calculator cannot be taught in a week, you need to be using it all year long).

Our school is philosophically opposed to screens in front of students (especially screens connected to the internet), so we only have one small set of Raspberry Pis that I use to teach computer science. I originally tried to introduce kids via Python on their laptops, but it’s much harder to do in a classroom setting without ironclad configuration control over the machines and environments. Fun times when half the kids have Python 3, half 2.7, half Windows, half OS X, and some have a borked install of Python 2.7 *and* 3 that sometimes works and sometimes doesn’t. You ask kids to install a nice editor and find out that half of them don’t have admin privs on their machines. Every minute I’m fighting that I’m not teaching.

I do break our class set of Pis out in upper level math classes to do things like Gauss Jordan elimination in Python. I think that people who are used to programming have a hard time perceiving how intimidating and difficult programming is to total novices.

The barriers to access even of a Python scripting environment as compared to TI-BASIC are huge.

The python documentation is written for experienced programmers, so the commands in Python aren’t as discoverable as the TI-BASIC commands, which are all happily listed in menus on the device itself.

Even the TI REPL loop is tighter than Pythons. Forget a colon at the end of a conditional, and the error you get is confusing. Whatever bad thing you do in a TI program, it just takes you to the line that is bad.

Beyond that, the syntax of BASIC is so compact and Python’s handling of math is so counter-intuitive. You don’t run into the same sort of actual computer science problems in Basic that you do with Python. Explaining to a classroom of children that 3/4 == 0, but that 3/4. == 0.75 because type systems is not a good use of (math) class time.

This is one I’ve ground my teeth on for the past five years, and nothing is better than TI-BASIC yet. I wish something were so we could ditch these expensive monstrosities. If there were and exact emulator (including a push-button editor) of TI-BASIC in the Desmos website, we’d be one step closer to ditching the calculators.

Yeah, fair point. I’m definitely not going to fight you too hard on my conflict of interest here. (Which I disclosed in the post!)

I only want to point out that dismantling TI’s monopoly doesn’t mean Desmos inherits it.

We have warmed states and other groups up to the idea of a web-based calculator embedded on their state exams, which only makes it easier for other web-based calculators to compete in the same space.

And I think your analysis of the economics here is largely right. Desmos doesn’t charge students or families but it does charge people who charge students and families, even if indirectly through school taxes.

But I’ve gotta point out quickly that we aren’t inheriting TI’s monopoly and we aren’t inheriting TI’s per-student prices. A TI-84 costs > $100 per student. Without undermining our business development team too much, I’ll say that our per-student costs are just a sliver of TI’s.