Do you know what Walt Disney’s first private venture was called? “Iwerks-Disney Commercial Artists.” He started the enterprise with his best friend Ub Iwerks. And it lasted a month.

Disney’s second partnership was started with Ub Iwerks as well. And when that failed and he moved to Hollywood from Kansas, he asked his friend to follow suit. Because Disney recognized the talent that Iwerks possessed.

Today, Walt Disney is a house hold name. No one knows the name of Ub Iwerks. Even though it was Iwerks who mainly refined Disney’s rough idea and created Mickey Mouse. It was Iwerks who drew 700 drawings a day to create the first Mickey animation in 2 weeks – a feat that others would have taken months for.

And yet, today, he is largely forgotten? Why?

The hill climbing problem

Computer science has this classic hill climbing problem. You are dropped at a random spot on a map. How do you figure out how to climb the tallest hill on the map every time?

The solution is easy if you can see the whole map from the top. But what do you do if you can’t see the whole map, and don’t even know how many hills there are? 

The easiest algorithm students develop is: take a step forward in the direction where land is higher. But this solution entraps you. You end up reaching the top of the hill closest to you. 

Its called the local maxima trap.

When you get to the top of any hill, the only way to go to a higher hill is by going down first. 

Iwerks vs Disney

Iwerks and Disney were partners, but when their second partnership failed, Iwerks focused solely on mastering animation. Disney, on the other hand, sought higher hills: setting up studios, selling animations, creating merchandise, building theme parks, and making movies. 

While Iwerks contributed significantly to creating Mickey Mouse, Disney earned more money and recognition by continuously pursuing new opportunities and expanding his vision.

The local maxima

Many of us fall in the local maxima trap. Perfecting our skills but in a limited scope. Only if we could see things from the top, we would realize that it’s the wrong hill that we’ve climbed. 

Because you can only climb higher hills if you’re willing to get down from your current peak.

The curse of low ambition

Half the people in an experiment are asked the question: Did Gandhi live beyond age 140? If not, how old was Gandhi when he died?

The other half is asked: Did Gandhi live beyond age 9? How old was he when he died?

The average answer for first half is 67. The average for second half is 50.

Isn’t that crazy? We are bound by initial information. 

• We get anchored by proximity. 
• We get anchored by things most familiar to us. 
• Early wins and losses guide our entire lives.
• We get addicted to the comfort of being at the top.
• We optimize for efficiency.

We give up on finding higher hills because we prefer predictible known comfort. Are you willing to go downhill first?

Escaping the local maxima traps

Do you know how the computer science students solve the problem of being trapped by the local maxima? How do you improve the chances of finding the highest hill when you don’t even know how many hills exist on the map?

1. Restart the algo

The first step most of them take is simply restarting multiple times. By trying different initial starting points, you can explore multiple paths and improve your odds of finding the highest hill.

And restarting is indeed a great algorithm. Its what Vince Papale did. A highschool teacher, he tried out for Philadelphia Eagles football team’s open tryouts at the age of 28. And became a professional footballer without having played the sport in college!

Papale was a great young athlete who played football, basketball, and track and field as a child. But he focused more on track and field because it gave him the highest chance of scholarships. He became the king of a small hill.

Only later in life, did he restart and find a higher hill.

2. Simulated annealing algorithm

Simulated annealing introduces randomness to the algorithm. Make random, less optimal moves to explore more of the space on the map.

Howard Schulz worked as a director of sales for Hammarplast: a Swedish kitchenware company. When he noticed a company placing unusually large orders for drip coffee makers from Hammarplast, he was intrigued. And flew to Seattle to investigate. Where he promptly switched jobs and became the head of retail at a much smaller Starbucks.

Going to Milan, Italy on a buying trip was a real turning point. He was inspired by the Italian coffee culture and forced his bosses to sell espresso beverages in America. He eventually bought the company and took it public – making it grow into the largest coffee chain in the world!

3. Mutation algorithm

A lot of the times, simply adding randomness to the algorithm doesn’t work well. Very often the algo would randomly take you to only one or two parts of the map. What worked well to fix this is taking a page from genetics and mutation. 

Introduce random changes. But maintain diversity in the changes.

Take the case of young Stewart. Started out in college majoring in chemistry. But shifted to history, and later architectural history. 

When that didn’t lead to a good job, became a stock broker.

Randomly, 10 years later, started a catering business from the basement. While catering for a book release party, a publisher gets impressed with the food and hosting skills and offers a book deal right away. And thats how Martha Stewart finds her right hill: with a ghostwritten cookbook.

Action Summary:

• Take a pause and reflect on your current situation. Are you stuck in a local maxima? Are you repeating the same things since forever because you like the comfort and the predictibility?

• If so, switch your algorithm. Restart. Try random things. Maintain diversity in your life and interests. Achieve your true potential.