Hi ,
"Let’s say you’re trying to teach a monkey how to recite Shakespeare while on a pedestal. How should you allocate your time and money between training the monkey and building the pedestal? The right answer, of course, is to spend zero time thinking about the pedestal." - Astro Teller
When you think about resource allocation in terms of monkeys and pedestals, it seems obvious to focus only on the monkey. After all, if you can't teach the monkey Shakespeare, then building the pedestal is a waste of
time.
Yet, every day in our own lives, we encounter similar choices, and many of us choose to build the pedestal instead.
Why?
Because it's human
nature to tackle the easy tasks first.
For data scientists, such a choice arises when building models.
One of the hardest (yet, most essential) aspects of data science is deploying models into production, with multiple surveys showing over 80% of ML models fail to
deploy.
Nevertheless, few data scientists even consider model deployment until the end.
For data scientists, ML models are pedestals and deployment is the monkey.
Tackle the monkey first and you have a greater chance of success.
Talk again soon,
Dr Genevieve Hayes.
