Abandon objectives and just get the job done

It was early morning when I got my weekly blog from my business coach. As my eyes scanned the first few lines, my mind began to rebel at what I was reading, instead of accepting it. A quiet “no” filled my mind… then a realization: I was disagreeing with my coach. The focus of the article was on goals, intentions and how to accomplish them using simple, daily steps as opposed to taking the time to develop a more complex strategy. My coach’s suggestion was that a person needs to act quickly by taking small steps towards a goal and then re-evaluate the progress towards achieving that goal along the way. It is only by moving forward that a person can discover the information that will help steer them in the right direction. Waiting for the path to reveal itself causes a significant delay in growth.

Should we try to make constant progress and reevaluate every step along the way?

Let’s think about it for a moment using a maze: Step into a maze with it’s very tall, imposing walls. Begin to move forward, problem solving and strategizing with each step that you take. You may run into roadblocks, get lost or even have to retrace your steps. You have to be creative, think on your feet and consider that your first idea may not be the right one to get you to the end. Is evaluating how far the exit is from your current location helpful? Will it help you get closer to the solution more quickly?

In the case of a maze, the answer is no. A large enough maze can elude someone for a very long time before they find the end. The reality is, most problems in life that are worth solving are much more complex than finding the way out of a maze.

Machine learning likes to define a “fitness” function as an expression of progress towards an objective in the search space. This forces the algorithm to act as an objective function. Let’s visit the maze again. This time with the player being a robot. The robot chooses a direction based on how close it can get to the end goal. Through deception, such objective functions may prevent the objective from being reached. See the maze on the right, the robot enters at #1, the end goal is #3, and the robot gets stuck at #2. While trying to reach the #3 the robot will always get stuck in a location that is close, but any further movement requires the robot to move away from the center which fails the “fitness” function and causes the robot to stop.

While methods exist to mitigate deception, they leave the underlying pathology untreated: “Objective functions themselves may actively misdirect search towards dead ends.[1]”

The realization: I need a better plan, or a better strategy.

The objective is to solve a maze of goals: going from point A to point B knowing that without a good strategy, it will be challenging to reach the destination .

In general, for our discussion here, let’s assume that ‘problems worth solving’ are seldom simple nor straight forward, and the majority of simple-looking-problems are most likely deceptive. Not realizing a deceptive problem for what it is can be costly.

Can we be spiritual and achieve our goals at the same time?

Eastern spirituality avoids attachment to objectives. While it sounds simple, by itself it is an objective, thus creating a circular dependency. To “abandon all objectives” seems to be a very confusing and random behavior that feels like there is no control over achieving goals. It is challenging for humans to comprehend random. In machine learning, this may be simpler. There is no personal investment in machine learning, therefore if a goal is not achieved it’s okay! This makes it simpler to take risks and try something new — abandon all objectives and see what happens!

Mathematically speaking: To abandon our goals, we need to simply abandon the use of the goal in our fitness function. How does this work? Let’s return to the maze one more time:

Human beings normally adopt the idea to explore the unknown. A robot can do the same, although the goal is to reach the center, let’s ask the robot to continue moving, but always to a place which it has never been before. This is a strategy that does not include the objective as a directive, but it will result in the robot reaching the goal — it may take a long time for the robot to reach the end goal, but the goal will be reached.

To improve the above strategy, machine learning deploys multiple agents/robots at once and requires each agent/robot to be unique, e.g. always go to where the other agents/robots did not. This allows for a much faster convergence on a correct solution, and it does not include the goal as part of the heuristic. Abandoning goals does not mean there isn’t a strategy or heuristic that leads to a solution. Defining this heuristic is sometimes very complicated and may require time to figure out.

A trickier example is the use of networking to achieve sales goals. The straight forward idea for a salesperson is to network to identify potential clients which could help him/her reach his/her sales goal(s). The salesperson is a ‘go getter,’ while some organizations choose to be ‘go givers’. Those organizations are using a different heuristic. Instead of networking to achieve the business goals, the networking is centered on referring other people and creating business opportunity for others. This approach is very different. It does not include the objective as part of the heuristic but at the same time strengthens relationships and leads to new business creation.

Here’s the bottom line:

Abandoning objectives doesn’t necessarily imply that goals will not be reached, it only forces one to have a heuristic approach that makes every small step along the way easy to make.

For myself I rather choose a heuristic to represent ‘who I am’ and consider ‘what I leave behind when it is all said and done’.

When I founded Multi-Innovation in late 2010, I used this idea to get projects. That’ right — I used giving to get. Today, after many years this is one of the base attributes of my team members that we love to give. It never ceases to surprise me what then comes back as a result of that giving.

[1] Abandoning Objectives: Evolution through the Search for Novelty Alone, Joel Lehman and Kenneth O. Stanley, Evolutionary Computation journal, (19):2, pages 189–223, Cambridge, MA: MIT Press, 2011