Humpty-Dumpty told Alice, “When I use a word, it means just what I choose it to mean – neither more nor less.”
Each of us, then, has our own Humpty-Dumpty that tells us what words mean to us and greatly influences our opinions and actions.
In the new field of data science, the practitioners are known as data scientists. This appears quite consistent.
In manufacturing, consider the plant manager, the person responsible for the enterprise’s cash inflow. The plant manager’s job is to keep production lines running so that product gets shipped to customers. If production lines stop, cash flow stops.
Before we consider what a plant manager is thinking about a machine learning project, let’s examine the manager’s “Humpty-Dumpty” just a bit.
A scientist is someone who does research. They develop a hypothesis or theory, run experiments to try to prove their theory, and if successful, write a research paper before moving on to a new and different theory. The scientist has at best a 50-50 chance of success.
An engineer is someone who takes requirements, selects components, follows procedures and processes, run tests, and refines the design to create something that has not existed before. The engineer has perhaps an 80% chance of succeeding except that running over budget and taking longer than originally schedule is quite common.
A technician is someone who takes an existing device or system, installs it, makes settings according to a prescribed procedure, and has it up and running typically within budget and on time. The technician is usually successful 98% of the time.
Now, let’s suppose this plant manager is contemplating a machine learning project, possibly one that centers around imaging. Do they want it implemented by a data scientist, data engineer, or data technician? If you were the plant manager responsible for cash flow, which would you choose?
