Maybe the best gift you can give your children is the experience of failure. It sounds radical, perhaps even a little uncomfortable, particularly when so much of parenting involves trying to protect children from frustration and disappointment. Yet by stepping in every time something becomes difficult, we may also be taking away opportunities for them to develop the very skills they will need to navigate an increasingly unpredictable world.
The question has become more pressing as artificial intelligence changes the way we think about work. In an earlier Trending Forward conversation, I explored what happens when AI begins taking over tasks and skills that once required years of training. The challenge isn’t simply predicting which jobs might disappear. It is figuring out how to prepare people for a labor market where the skills that are valuable today may not look quite so valuable a decade or two from now.
Children growing up today will enter that world without us knowing exactly what it will look like. Some of the jobs they eventually pursue may not exist yet. Two decades ago, working for a YouTube creator would hardly have been an obvious career aspiration; today, it is an entirely plausible profession. Rather than trying to guess the exact careers that will emerge, it makes more sense to focus on the human capabilities that can travel from one situation to another: curiosity, judgment, problem-solving and perseverance.
Those ideas were at the heart of my latest Trending Forward conversation with Rachele Harmuth, Chief Product Officer at CrunchLabs, the education and entertainment company founded by former NASA engineer and YouTube creator Mark Rober. I came into the interview already familiar with the company’s approach. Years ago, I bought a CrunchLabs Build Box subscription for one of my granddaughters and spent hours assembling projects with her. The projects were clever and genuinely educational, but what stood out was that they did not always go together perfectly on the first try.
Looking back, I suspect the failed attempts were at least as valuable as the finished projects.
AI changes what kids need to learn
For generations, education has placed considerable emphasis on accumulating information and mastering established skills. Children learn dates, formulas, languages, coding and other disciplines with the expectation that those abilities will help prepare them for work later in life. Knowledge still matters, but AI is making it increasingly easy for machines to perform many tasks that once depended on a person’s ability to recall information or execute a process correctly.
The more difficult question is what happens when there isn’t a clear process to follow.
Harmuth and I talked about the difficulty of predicting which careers will thrive 15 or 20 years from now. The answer isn’t simply to identify today’s most promising technical skill and teach it to every child. By the time today’s eight-year-olds enter the workforce, the technology around them will almost certainly have changed. The better investment may be in teaching them how to approach unfamiliar problems, work through uncertainty and keep going when the first solution doesn’t work.
CrunchLabs summarizes that philosophy in a simple phrase: “Think like an engineer.” It isn’t a suggestion that every child who builds one of its projects should eventually design bridges, develop robots or work for NASA. Engineering, in this context, is a mindset for solving problems: define what you are trying to accomplish, brainstorm possible solutions, test one, understand what went wrong and make an adjustment.
Curiosity, creative confidence and an ability to embrace failure sit at the center of that approach. The first two are relatively easy for adults to understand. The third is much harder.
Making failure less threatening
Parents and grandparents naturally want to fix things. When we watch a child struggle with a puzzle, an assembly step or a new skill, the instinct can be almost automatic: move that piece, turn it around, let me show you. Sometimes intervention is necessary, but removing every obstacle also removes the chance to discover what happens when you work through frustration yourself.
Harmuth made an interesting comparison during our conversation. Kids already understand this process when they play video games. Nobody expects to pick up a controller and beat a difficult game on the first attempt. You fail, restart, adjust your approach and try again. A failed attempt isn’t evidence that you are incapable of playing the game. In many ways, it is how you learn to play it.
The physical world often produces a very different reaction. A drawing doesn’t turn out as expected and a child decides they are bad at art. An instrument squeaks and suddenly they aren’t musical. A build fails and the conclusion becomes, “I can’t do this.” The problem isn’t the failed attempt itself; it is the meaning we attach to it.
Mark Rober’s work offers a useful counterexample. His videos showcase remarkable inventions, but the finished creation doesn’t tell the whole story. Behind the polished result are failed ideas, prototypes and iterations that didn’t work. The willingness to show those failures is part of what makes his approach to engineering accessible to children in the first place.
CrunchLabs applies the same philosophy across its products. Build Boxes introduce younger children to hands-on physics and engineering projects, while the Hack Pack gives older kids an opportunity to explore robotics. The company’s CrunchLabs Mysteries books take the idea into another format, turning reading into a hands-on STEM adventure and asking children to solve puzzles and build devices themselves. I recently looked at that expansion into storytelling, and it reflects the same underlying idea: STEM doesn’t have to mean sitting down and being told what to learn. It can involve figuring something out for yourself.
The company is also expanding into retail, summer camps and Class CrunchLabs, a nonprofit initiative offering free STEM curricula for grades 3 through 8. The formats are different, but the lesson remains remarkably consistent: try, fail, learn, adjust and try again.
Resilience Is a Teachable Skill
My favorite moment in our discussion was Harmuth’s perspective on resilience. We often treat resilience as an innate trait—you either have it or you don’t. Harmuth argues it is actually a muscle built from specific skills: problem-solving, perseverance and adaptability.
Parents can help build those skills without deliberately making childhood harder. A child working on a skateboard trick, learning an instrument, writing a story, experimenting in the kitchen or building a backyard fort is doing much of the same mental heavy lifting as a child working through an engineering project. Something doesn’t work, so they have to figure out why, make an adjustment and decide what to try next.
Curiosity gets a child wondering what might happen if they try something different. Creative confidence gives them the belief that they can figure it out. When the first attempt fails, resilience helps them move past “I can’t do this” and toward the much more useful question: “What should I try next?”
Those habits could become especially valuable as AI becomes better at generating answers. Humans will still need to decide which questions are worth asking, evaluate whether an answer makes sense and work out what to do when the problem in front of them doesn’t have an obvious solution.
Skills for a Future We Can’t Predict
None of us can say with any confidence which careers will thrive 15 or 20 years from now. Many of the jobs today’s children eventually pursue may not even exist yet. Two decades ago, the idea of working for a YouTube creator as a dream job would have sounded nonsensical. Today, it is an entirely plausible career path.
Preparing children for one particular version of the future may therefore be less useful than giving them skills they can carry into whatever that future looks like. Curiosity, judgment, problem-solving and perseverance aren’t tied to a particular profession. They can serve a child who becomes an engineer, an artist, a teacher or something we haven’t even thought to call a career yet.
“Think like an engineer” works as a philosophy precisely because it isn’t really about engineering. The process is simple enough: figure out what you are trying to accomplish, brainstorm possible solutions, test one, see what went wrong and adjust. A child doesn’t have to know anything about engineering to benefit from learning how to approach a problem this way.
Years after buying that CrunchLabs Build Box for my granddaughter, I think differently about the projects we built together. The satisfaction of getting something to work was certainly part of the fun, but the moments when we had to stop and figure out why something wasn’t working may have been more important. Getting the answer right was only part of the experience. Learning how to get there mattered too.
Parents naturally want to fix things for their children. We see frustration and want to make it disappear. Sometimes helping is exactly what a child needs; other times, giving them the space to struggle with a problem for a little longer may be the more valuable response.
The world today’s children inherit will be shaped by technologies we cannot fully anticipate, and the jobs they eventually do may look very different from the ones we know today. Preparing them for that uncertainty doesn’t necessarily mean teaching them every new technology as it arrives. It may mean giving them the confidence to face something unfamiliar, the resilience to fail without losing heart and the curiosity to keep experimenting.
My conversation with Rachele Harmuth left me with a simple thought: we may not be able to prepare our children for every problem they will encounter, but we can help them become better at working through problems. Even if they never become engineers, learning to think like one could prove to be one of the most useful things we give them.
