Call Me Professor James

I wrote a memo a while back called The Calculator Question, about how I’m approaching education with my 7 year old daughter. Well, I’ve decided to put my money where my mouth is and test some of those ideas in person. This fall I’m teaching Innovation and Entrepreneurship at UT Dallas… one section of undergrads, Tuesdays and Thursdays at four. I said yes almost immediately and I’ve spent the past few weeks figuring out how I want to run it. Nobody is calling me Professor Wang, by the way. I’ve heard way too many Wang jokes in my life.

History Rhymes

The policy I’ve landed on is that my students can use AI for everything. I want them to use it. Aggressively. I haven’t come across a startup in my own deal flow that isn’t using these tools for everything from generating backend infrastructure to the pitch deck itself. Keeping students away from these tools would prepare them for a working world that no longer exists.

I was in school when the internet stopped being optional, and my teachers spent a few genuinely confused years deciding whether we were allowed to use Google. Wikipedia was treated as somewhere between laziness and cheating. Nobody says that now. What actually won with Wikipedia was procedure, not permission. Start there, never cite it, follow the footnotes down to the primary source. The professors banning it got the tool wrong, but they were right about the risk. Their concern made sense, because students were taking someone else’s finished answer and presenting it as proof that they had done the thinking themselves.

The comparison breaks in one major place. Google hands you sources and leaves the judgment to you. You still have to figure out which link is credible, who funded the study, what the other side would say. AI hands you the conclusion with the judgment already performed. The reasoning is exactly the part that got cut out, and the reasoning is the whole thing I’m supposedly there to teach. Students using the tool to do the work doesn’t bother me at all. The challenge is figuring out how to test how they think instead of what they can produce.

What Actually Changed

A market map, a comp analysis, a five-year model, a twelve-slide deck. Every one of those used to cost a student a weekend, and the cost was paid in manual labor, because you couldn’t produce the output without absorbing the material on the way. Handing in the final product was a decent proxy for the absorbing. Production is now roughly free, so the assignment itself has stopped telling me anything about the person who turned it in.

This is the same point I’d make about starting a company today, and I’ll probably make it in class twenty times before Thanksgiving. Building an MVP has never been cheaper. One capable founder with the current toolset ships in a weekend what took a seed-funded team six months in 2015, which is wonderful, and which also means an MVP no longer distinguishes you from anyone, because everyone has one. The bottleneck moved. It’s distribution now, getting a stranger to care and try and pay and stick around, and that was always the expensive part… it just used to be hidden behind the other expensive part. The same thing happened to my assignments. Once the production cost goes to zero, what’s left over is judgment, and judgment only becomes visible when someone has to defend a position out loud without help.

How The Grade Works

So half my grade is live performance in a room with no laptop, split between cold-call participation and presentations. I’m going to run the room on the Socratic method, which in practice means calling on whoever looks sleepy. Presentations are five minutes to pitch and ten minutes of questions.

That two-to-one ratio of interrogation to pitch is deliberate and it’s inverted from how these things normally run. Most classes and most pitch competitions give you fifteen minutes of deck and three polite questions at the end, which trains students to build beautiful decks. I’ve passed on companies with genuinely lovely decks because I asked where a CAC number came from and watched the founder stare back blankly. I’ve also been on the other side of that, sitting across from someone who asked me a question that I should have been able to answer and couldn’t. Ten minutes of hostile questions is the actual job, and I didn’t invent it as a classroom exercise.

Use whatever you want to research, model, design, and rehearse, and then stand up and defend it cold while the room tries to knock it over. Getting to a number faster just means you’ve bought yourself more time to understand the number, which you will need, because I’m going to ask what happens to it when churn doubles. Socrates never collected an essay in his life, and the oldest teaching method available turns out to be the best answer to a machine that writes any document on demand. It doesn’t scale past about forty students, which is roughly what I have.

How The Final Pitch Is Judged

The final presentation is the part I’m looking forward to most. I’m bringing in outside judges, investors and operators who do this for a living. Each judge gets a million dollars of imaginary capital and can put up to five hundred thousand into any one team, three teams maximum. They can also write smaller checks, or none at all. Allocations get written down independently before anyone compares notes, so nobody gets to follow the room, and then each judge says out loud where the money went. A judge who funds nobody has to stand up and explain that too. The top of the leaderboard picks up five bonus points on the presentation grade, scaling down from there by how much each team actually raised.

A few things I like about that. Eight teams and a three-team cap means every judge passes on at least five of them, and passing is most of what this job actually is, so the room watches it happen live instead of hearing me describe it. The feedback comes from people with no relationship to these students and no particular reason to be gentle. And the semester ends with a number attached to the work, which is what happens to founders eventually whether they asked for it or not.

I’ve spent a long time on the evaluating side of this and none at all on the teaching side, so I don’t know exactly how it’ll go. I’ll find out in December alongside everybody else. It should be fun.