Hi ,
In my last email, I asked readers what they considered to be the opposite of success. Here are some of the answers I
received:
- "Complacency is when you were successful. It may not be the polar opposite but it is in opposition." - Paul Rayburn
- "Having everything you "want" but feeling empty." - James Turner
- "Stagnation...not being able to do, think, play, work on 'what could be'." - Rod Aparicio (and please check out his full response to this question HERE)
- "Mediocrity... Muddling along, kinda doing ok, but not having (an) impact (or) giving (your) team members the opportunity to see how much impact they can have." - Jonathan Dursi
As a data scientist, with experience with reinforcement learning, I see the opposite of success slightly differently. To me, the opposite of success is
exploration.
Reinforcement learning is about learning through experience. To gain this experience, the training of a reinforcement learning "agent" involves two types of actions: exploration and exploitation.
Exploration involves the agent interacting with the environment in which it exists to gain knowledge - even if that knowledge is of what doesn't work.
Exploitation, on
the other hand, involves making use of the knowledge gained through exploration to maximise reward.
And as with just about everything in data science, a trade-off exists between the two. An agent can either explore its environment or exploit it, but never both at once.
If trained correctly, therefore, a reinforcement learning agent should succeed when exploiting its environment, making exploration the opposite of success.
Here's the thing...
Humans are hard-wired to be loss-averse, which means none of us like to fail. Yet, ironically, the only way we can succeed and achieve our goals is by learning what failure involves and how to avoid it.
But if failure is necessary to succeed, is it truly failure?
Reframing failure as exploration makes it easier to accept - and
more likely you will ultimately succeed.
Talk again soon,
Dr Genevieve Hayes.
p.s. I recently had the opportunity to speak to reinforcement learning expert Prof. Michael Littman on Value Driven Data Science. You can listen to the episode HERE.
