Why K-12 Education is Key to American Leadership in AI

This week the Trump administration released an executive order on American leadership in artificial intelligence (AI). The order outlined education and funding priorities necessary for the U.S. to remain competitive in AI, one of the most rapidly advancing technologies in history. Recently Iridescent CEO and founder, Tara Chklovski, shared her initial thoughts about the plan with Education Week. While it’s great to see the administration prioritizing AI research and workforce retraining, she noted it’s missing two crucial pieces: K-12 AI education and ethical development of AI.

“The key to U.S. competitiveness in AI may be locked inside the minds of the children and teenagers who will grow up in a world increasingly defined by automation technologies,” Chklovski explained. “Without a concerted effort to teach AI principles to children, the U.S. risks putting students at a disadvantage once they enter the global workforce.”

A family works together to create their own “self-driving car game” as part of the AI Family Challenge.

K-12 AI education must go beyond technical information in textbooks

Highlighting countries making substantial investments in AI education like China, she pointed out hands-on, project-based curriculum as an opportunity for the U.S. to create richer learning environments. Teaching soft skills alongside technical ones helps prepare learners for a career path where the impact of emerging technologies on the future of work is less known.

“Countries like China, with students frequently outperforming American students in science and math, are investing a lot of money in AI education. The U.S. has an opportunity to excel by building skills that go beyond textbooks. One way is connecting technical skill building with opportunities to solve real-world problems. Through our AI Family Challenge program we’ve found that challenging children and adults to learn about technologies like AI and then having them apply those skills to solve  real-world problems helps them build job skills like curiosity, creativity, and collaboration.”

A family works together to complete a project as part of the AI Family Challenge.

The ethical development of AI needs the same level of care and attention as privacy concerns

In addition to K-12 AI education, the plan doesn’t address the issue of ethics. Trust and safety considerations like data privacy are important. But issues of bias, fairness and algorithmic transparency are also crucial to ensure AI technologies are representative of the populations they serve.

“There is an alarming lack of diversity among the people who are currently building the algorithms transforming industries,” said Chklovski. “If the issue is not addressed on a national scale, the gap between the people who can access and provide input on building the future of AI and those who cannot could lead to long-term bias against the latter and greater economic disparity in the country.”

Adults and children must feel empowered to learn about new technologies and have the opportunity to use them in meaningful ways. The new executive order puts attention on an important conversation. But, it is only the beginning of what needs to be a much larger, ongoing partnership between government, industry and academia.

Persevering Through the Unknown: My Conversation on Emotion AI and Problem Solving with Emily Mower Provost

Tara Chklovski: What inspires you? Emily Mower Provost: The idea that we can use technology to understand people. Technology gives us new ways to learn about how humans communicate. It allows us to tackle the unknown. Advances in technology allow us to focus on problems that were previously deemed too difficult to handle. For example, […]

Preparing children for the 21st century workforce

By Katy Santa Maria Families sat quietly in the cafeteria, awaiting instructions. They were unsure of what to expect. When Curiosity Machine Family Science sessions began at Los Angeles’ 10th and 20th Street Elementary Schools in the Spring, it was up to University of Southern California engineering students, along with parent leaders, to build the […]

The Four Freedoms of Play and Common Core Standardized Testing

I was inspired by Scot Osterweil’s recent presentation at GLS, in which he presented his four freedoms of play:

  • Freedom to Experiment
  • Freedom to Fail
  • Freedom to Try on Identifies
  • Freedom of Effort
What makes this framework most interesting is that to Scot, these are not only the four freedoms of play, but also the four freedoms of learning. Good learning environments also need to contain these freedoms to be effective, an idea very much resonant with James Gee’s view on games and learning.

Although these freedoms are not a particularly new idea, it was new to me this year, and it really helped crystalize several previous thoughts I’ve written about. It definitely resonated with my ideas that agency is not a binary quality of an activity, but that learning activities can contain different degrees of agency. This framework helped illuminate some of those different degrees to which an activity can contain freedom/agency. 

But most interestingly, Scot noted in his talk how school doesn’t contain these freedoms, despite the fact that both games and learning do.  His challenge to us was to imagine a school environment that did contain these freedoms. I found it especially interesting to think about in relation to Common Core and standardized testing, which led to the question of this blog post: could a school environment that was constrained by standards ever achieve these four freedoms? Let’s try to break down the freedoms and answer that question.

1. Freedom to Experiment

The freedom to try out variations and play around with a system.

So, how do standards rack up to freedom number 1? Ok at best, I suppose: it could be worse.

The standards themselves are quite open and emphasize higher-order skills, which can be learned in many different contexts. This provides some grounds for exploration and experimentation in learning. At the same time, it doesn’t require such experimentation and teachers can use a very structured, linear lesson plan to teach whatever standard it is they are trying to teach. So, it could go either way here.

The standardized tests do ruin the freedom to experiment a little bit, as ultimately you can do as much experimenting you want until you get to the tests, and then you need to learn how to pass tests. But hey, it could be worse–with a good teacher, you can still spend most of the year learning higher order skills through experimenting.

2. Freedom to Fail

The freedom to fail as often and as much as one would like, without consequences.

How do the standards rank to freedom number 2? Definitely a big fat zero.

The standardized tests very explicitly do not allow for failure–you need to get them right, and get them right that first time, otherwise there are consequences for both you as a student and your teacher. Any system that has high stakes tests by definition does not include freedom to fail; that is literally a part of the definition of “high stakes.”

3. Freedom to Try on Identities

The freedom to engage in an activity with the role, identity and perspective of your choice, and to change roles as you wish.

C’mon standards, how about freedom number 3? Another big disappointment.

The whole point of a standard is that by definition, everyone fits the standard and looks the same. A standard creates one identity that everyone has to mold into by the end of each year- your ability to try on different identities, to specialize in your learning in any way is irrelevant and not encouraged by the standards. In fact, the standards are very explicitly reinforcing uniformity, the exact opposite of this freedom.

4. Freedom of Effort

The freedom to try as hard or as little as you like.

What about freedom number 4? Well, standards probably didn’t make school any worse off at this than it was before, but there’s certainly no improvement here.

Can you choose to try hard in your learning one day, and not in the next? I’m not sure this goes against the standards as much as it goes against our very idea of school, which is comprised of kids sitting in chairs and being forced to put effort into listening to a teacher at prescribed times. The standards certainly don’t help the issue (there’s things that need to be learned this year by everyone, whether you want to or not!). But school itself was structured quite heavily against this freedom well before the standards arrived.

The standardized tests themselves, of course, make everything that much worse. Testing days, which become more and more extensive every year, force students to sit and expend effort on challenging tests hour after hour. Certainly there’s no freedom of effort here, you have to show up and you have to take the test and do well. 

Conclusion: School will never embody the four Freedoms of Play as long as it has standards regulating what should be taught.

Of course, I’m sure this conclusion has some of you shrugging your shoulders, saying “So what? School shouldn’t feel like play, what’s the big deal if standards are promoting that atmosphere?” To which I’d point you back towards Scot Osterweil’s talk, because these Freedoms of Play are not just for play and games, but are also the freedoms that embody good learning. What does it say about schools when standardized testing renders schools incapable of being good learning environments?

5 Signals that an “Educational Game” Isn’t Really a Game

Kids love games, but why do they hate educational games? The short answer is that most aren’t truly games, because being gamelike means a lot more than having flashy graphics and a point system. As an educational game developer, I think one of the most damaging aspects to this industry is when people call things games in order to get kids to play them when they clearly aren’t games.

How can you spot the fake games masquerading as educational games? Here are a few signals I’ve picked up on over time.

1. When walking through a demo of the game, the game designer stops to say “And this part is where the learning occurs.”

The learning should be everywhere, not in one part of the game. If you can compartmentalize the part of the game that is about learning, you did something wrong. One such example would be breaking up the game to show a player an instructional video- if you are using a video to teach, then you are not using the gameplay to teach.

It is ok for games to be more or less explicit at different points in the game in how they teach. Most teaching in games is pretty implicit, but often when first learning something new in the game, a more explicit description of that new component is given. And of course, boss battles are very explicit tests of certain skills. So games can scaffold and foster learning in different ways throughout the game, but there should not be a distinct point in the game which is “the learning part.”

2. “And then to add the motivational element, we added a game component to the lesson.”

Is that all games are, a “motivational element?” Sounds a bit Skinnerian in it’s view of human nature–games just add a bunch of extrinsic motivators to make something feel “fun” and motivate someone to play. Of course, if you’ve ever played a meaningful, complex game, you know that games do far more than that, and I’ve discussed the role of intrinsic motivation in games elsewhere, so I’ll stop commenting here.

3. Excessive use of the word “fun” in describing why the game works.

This is a bit similar to the last signal, but there’s something peculiar about the word “fun” to me. Fun is rather illusive to describe and define. As Raph Koster puts it, we can’t really define what fun is, but we know it when we see it. To me, fun is extremely hard to just add to something like a game. Fun is not a component of a game, it is an emergent property of a well designed game. Fun doesn’t make a game work well: a game that works well is fun. Some people talk about “fun” as this easy-to-grasp element that you just tape onto a learning activity to transform it like magic into a game. Those people always worry me a bit.

4. Extensive in-game tutorials, as videos or text.

Will Wright defines games as “a series of meaningful choices.” When you are reading or watching something, you are not making choices–you are being a passive learner, rather than an active one. If there’s one defining feature of how games act as educational tools, it’s their ability to foster active learning. At its core, games need to be active, and so I am wary of educational games that overly rely on the more traditional, passive forms of learning.

5. Multiple choice items in the game that have clear right answers.

This one is a bit complex, as having a player select a choice in a game is not a bad thing, in general. For example, World of Warcraft or most modern RPG’s with have choices that can result in divergent storylines. But in these cases, there is a clear “right” choice, just different storylines. And strategy games like Faster than Light can implement choices in a strategic way- some choices are riskier and produce better payoff than others, but no choice is necessarily right or wrong. It’s up to you to judge as a player whether the risk is worth the reward for you at that point in the game. Quandary is another example of an educational game with interesting choice-making- but notice the choices aren’t multiple choice, they have an interesting drag and drop system for figuring out the best way to interview someone. And the game is defined by it’s clear lack of a “right” answer (although there are right ways to argue for answer by using appropriate evidence).

The other complexity is that games that implement right-answer multiple choice items sometimes do so in a way that doesn’t seem like a multiple choice list. Think Math Blaster, where the right answer is shooting the right asteroid. Numbered asteroids may seem more “fun” than numbered lists but there’s still a list floating around in outer space, and one of those asteroid is “right.” Of course, even worse and more obvious are games that always show 4 items in a list, or even label them a, b, c, and d.

Do you have additional signals you’ve noticed that are worth adding? Tell us in the comments!

Should you pay students to go to school?


I just came back from the National Science Teachers Association (NSTA) conference in New Orleans. It was not as intense as my last year’s experience at Boston, but fun all the same. I met one very interesting Canadian teacher who teaches college students in Qatar.
She was saying that students there are all paid to go to school and college. Many of them come in Lamborghinis and the biggest problem is motivation. Qataris have a lot of money and outsource all their jobs to foreigners. So the government has started paying students to go to school and college and learn some basic skills.

I thought this was a fascinating philosophical problem: if you take money and financial stability out of the equation, what reason do we have to get educated? To flaunt our sophistication to others and thereby distinguish ourselves from the crowd? For intellectual pleasure? Would you still want an education if you were Robinson Crusoe? Is it to leave a legacy for someone? If you remove ego from the equation as well, would there be any defense left for getting an education?

Engineers as Teachers


How many of us have pursued a career because of a passionate mentor or teacher? I personally chose engineering mainly because my father shared his love for it with me.

Children in urban schools rarely come in contact with highly motivated, intense scientists or engineers. So when you go into their classrooms you change the way they think. You prove that they too can aspire to and pursue a career in science or engineering.

This experience was a very powerful for me. I decided to devote all my energies to motivating underserved children to develop important skills, work hard and aim high.

We need to build a society that can analyze situations, think objectively and make intelligent decisions. We have to change the world ourselves!

I started Iridescent, a non-profit organization dedicated to bringing the best and most exciting science experiences to underserved communities. The model is essentially to help engineers use their skills to develop interesting, hands-on science activities for children and to inspire them to pursue a career in science or engineering.

The Aerospace & Mechanical Engineering Department at USC has been a very strong champion of this program. The department offers a course for undergraduate students (AME 490) that enables them to earn 3 credits for developing and teaching hands-on science courses. Over the past year, we have had 13 students who have conducted 9 courses reaching more than 270 children and parents. (Some of the courses we conduct are for entire families!)

Each course is for ten hours and the course topics are all very unique and real-world. So far we have developed and taught courses on: Aeronautics, Biomechanics of Diving, Cardiovascular Mechanics, Heat transfer and Energy Efficient Houses, Physics of Sailing, Renewable Energy, Rockets, Science of Sporting Equipment and Structural Color.

Every instructor goes through a 16-hour, training program in which he/she learns about motivation, interest, educational psychology, inquiry-based science, lesson planning and evaluation. The instructor also learns how to make assessments for each session and ensure that the students are always learning.

This model of Engineers as Teachers is very powerful as it changes not only the school children
that are being taught, but also the instructors.

The biggest challenge is communicating complex and abstract concepts such as Reynolds number, diffraction, dual nature of light, gravity etc without using any mathematical equations or scientific terms. You have to assume zero prior knowledge (regardless of the participant’s age) and then build up block by block.

The other rewarding aspect is learning about the community and witnessing firsthand a parent’s drive to provide the best for his/her child. We have had parents work two night-shifts and come for a session first thing in the morning. Another parent drove all the way from Santa Barbara (to Los Angeles) every Wednesday night to attend the sessions. It is very inspiring and humbling to come in contact with people who may be less educated than you but may have more ambition, more drive and more determination.

The next step for us is to bring more engineers into the program. We have many more schools than we can provide courses for.

We need passionate, dedicated, visionary engineers and scientists to help inspire a new generation to solve the world’s biggest problems.

Let’s make it happen!