Teenagers out of School

at a camp for potential future entrepreneurs

The timing was unfortunate. A venture capital fund that was holding a three-day camp for teenagers invited me to try out my teach-a-thon idea for an hour and a half on the second day. That was October 7.

The VC’s were looking to attract young people with a passion for engineering and impatience with classroom learning.

The Kids

There were about 20 young attendees at the camp. They ranged in age from 14 to their early 20s. About half were girls. Almost half came from Canada, where the camp fit well with the work-study philosophy of the University of Waterloo and with a program for younger Canadians interested in STEM.

The kids had deep knowledge of particular areas of interest. Their knowledge was acquired out of the classroom.

One of them literally tries to build rockets in his family’s garage. He also was very thoughtful about the process of creative engineering itself. He said that you start with a set of problems, and then you whittle them away one at a time until you have something that works. He noted that he had not yet whittled away enough of the problems with his rocket to actually test it.

To me, some of the kids seemed too pleasant to become young entrepreneurs. Instead, I picture them being recruited by large organizations and quickly working their way up into management positions. In two or three decades, they might finally decided to start their own firms. But although I can see them getting along in large organizations, they are interested in learning, not in school.

A few of the kids combined their interest in STEM with an emotional need that I associate with starting a business: a drive to achieve fame and success; or a rebellious streak that would not fit in an established organization.

This was a first-time attempt by the VC fund to run a camp for teenagers. I thought that the logistics and spirit of the camp went remarkably well.

Labs in Garages

The first day (October 6), we took a “field trip” to visit six start-ups in STEM fields. All of them were located in industrial garages. All of them were very hands-on in what they were doing. To me, they looked like off-campus laboratories. Energy-related start-ups had lots of wires. Those that involved chemistry had glass tubes. A robotics firm had a room that looked like it contained parts from dozens of erector sets.

I was worried that none of the start-ups had an aggressive salesperson. They were using VC funds the way that another lab might use a grant from a foundation or from the government. If the strategy is to build the mousetrap first and then think about sales later, I have doubts about it.

I strongly prefer to see a company that is closely testing its product in the market, even if it is not yet built. You may find that you can tweak your initial idea to get revenue with something that is not as challenging to create. With revenue, you can get your dream product built faster than if you have to get by on a minimal budget doled out by funders. And VCs may be especially stingy right now, given the state of capital markets.

The Kids as Criminals

The second day, the kids were in action in two sessions. The first one was a hands-on lesson in lock-picking, led by a guy who has been giving such lessons for close to twenty years. Everyone in the session was given a set of locks and lock-picking tools. With some minimal teaching from the leader, soon all of us were picking locks. Later, we were given handcuffs and shown how to break out of them.

To say that the kids were wowed by this session would be an understatement. But by this time I had heard about the pogrom in Israel, and I did not have the focus to give the exercises the attention that they deserved.

I think that the fastest lock-picker was a young woman who when she spoke really made an effort to choose her words carefully. She interrupted herself often. Someone who might be brilliant but unable to ace an interview.

Chatbots as left brains and right brains

The session I led was probably a letdown after the lock-picking. I asked the kids to use chatbots to create a course outline or book table of contents on the topic of Booms and Busts in American financial history.

I had three goals: think about the role of the new AIs as teaching tools; get a feel for using the chatbots; and learn some American financial history.

I had them break up into teams. I let them form their own teams. As it turned out, one table of about 8 was all girls, one table of about 8 was all boys, and one small group was two boys and two girls.

The girls organized themselves cheerfully and diligently. About half the boys did not do the exercise at all. They futzed with their computers or with the handcuffs.

One of the best lock-pickers did not do my exercise at all. Based on his temperament, I wrote him up—as a real potential entrepreneur. I have a stereotype that an entrepreneur is rebellious.

What struck me about the Web in 1993 was its potential to allow someone like me to try an idea without having to wade through the permission swamp of a corporate bureaucracy. A tiny start-up could distribute content on the Web as easily as a media giant.

As a technology, I think that the new AI’s have as much potential as the Web to reshape how we do things. But as of now I am not seeing the same degree of opportunity for the little guy. If AI is going to lower the cost of entry in a large segment of the economy, I have not observed that yet.

After the exercise, I went around the room asking the kids what they had learned about the learning process. I was hoping for some higher-level thinking about what it would take to make the chatbots able to serve an independent learner. I was a bit disappointed in the responses, which were more focused on what they learned about writing prompts.

Next, I asked students to shift away from left-brain to right-brain thinking. I gave them an assignment to do something creative with chatbots, keeping with the topic of the history of booms and busts.

Interestingly, each of the three groups chose to have the chatbots compose rap poems. One of the teams ended up with something very close to Fear the Boom and Bust, but with different characters (Milton Friedman represented the right, I forget who represented the left).

As of now, I think that people over-estimate the left-brain skills (logic, reasoning) of chatbots, treating them as research assistants. The first exercise I gave was oriented too much in that direction.

I think that people under-estimate the chatbots’ right-brain skills (creativity, role-playing, synthesis). The second exercise was a mild illustration of right-brain skills.

But people constantly seem surprised at how well chatbots do at relating to people (e.g., in a medical setting they score well in bedside manner). People refuse to believe that chatbots can be warm and relatable. That side of the new AI is something to which you should pay attention, even if you find it offensive.

Long ago, I said that the future belongs to the autodidact, the person who can learn independently. The kids at this camp reinforced that view.