Failure

Creating And Producing

History

When I first started learning Python, I was somewhat entranced with the idea of machine learning.

I didn't really know what that term meant, but I thought it might be fun to find out.

Soon after learning the programming basics, I started looking into a library called Gensim, which uses the tag line "topic modeling for humans."

Although I didn't quite understand everything I was reading, ideas started swirling in my head.

I thought of a hypothetical tool that could take any sci-fi movie as a prompt and, as such, deliver the likeliest philosophical topics relevant to such a film. It might even point me to other movies or media within the same topical realm.

I had no idea how I would accomplish that, but the idea led me down some fun paths.

I learned to scrape the web to try and get as many movie reviews and synopses from as many sources as possible.

I found out about the OMDb API, this is before I even knew what an API even was!

That's also when I learned about the Internet Encyclopedia of Philosophy, which was fun to peruse (and scrape) for content.

My journeys also took me to Wikipedia's dump files, something I never would have realized even existed.

Also, while working with Gensim, I learned about vectors and transformers. I also learned how to differentiate between Python's iterators and generators, and when it might be a good idea to use the latter.

I even learned some very basic plotting with Bokeh and Seaborn and discovered Jupyter notebooks.

For much of that time, especially at the beginning, I barely knew what I was doing, but I learned (and subsequently forgot) a lot of stuff. But more than anything, I was consistently engaged. This was before I even learned about the Python community.

I was excited when I figured out a way I could provide a word or phrase into a notebook and hence generate a plot of the top ten most relevant philosophers/philosophical themes.

It was at that point that I thought I may want to create a web application in order to fully realize the dream!

It was ambitious...

...and I failed.

Eventually, I abandoned my interest in machine learning in lieu of playing around with web frameworks and figuring out how to build things on the web.

Revisionist History

Let's say... I was having this idea today.

What if I could link sci-fi movies to philosophical contexts, and create a recommendation app that used topical similarity to link different media to philosophical texts (and vice versa).

Today, I might not need to learn a programming language. I probably wouldn't need to learn about vectors, or generators, or iterators, or data dumps, or web scraping, or web frameworks, or API limits, or anything like that.

I could pay a subscription fee to some service, craft some prompts asking specifically for what I wanted, learn to iterate on clunky prototypes, rinse, repeat... and then I'd have an app that somewhat fulfilled my initial ambition.

That's cool, I suppose.

But I probably would not have gained any of the things described in my previous section.

My success would actually be a tremendously lost opportunity.

My point is, in my failure to build a recommendation app as I had once envisioned, I gained so much more.

I really don't think I would have had the boldness to submit proposals to speak at PyCon and other various conferences. I wouldn't have gained enough self-confidence to keep learning new skills, and I probably would not have had the courage to switch careers in my mid 40s.

Time

This, of course, is an anecdote. It doesn't prove anything.

It's just a long and fanciful way of demonstrating that aphorism attributed to Thomas Edison.

I have not failed. I've just found 10,000 ways that won't work.

When I think of creating something, I never ask myself, how quickly can I get this done so I can work on another thing?

For example, recently, I started sketching a drawing of a familiar face to me. I haven't seriously sketched for a very long time, but this time, my daughter asked me to draw something, and I decided not to cop out by doodling something quick and quirky.

My pencil strokes started soft and ethereal. I was using a photograph as a reference, one I know very well—it's what I use as my phone's wallpaper.

Slowly, the lines on a page became an apparition, and then a face began to emerge.

After adding just a touch of color, I decided to leave it as is. I think I feared I'd mess it up if I kept going.

The picture's not perfect. I had a hard time with the nose. The hand resting on the cheek was hard to reproduce. The upper lip is a bit off center.

Now, going back to the question that kicked off this reflection.

How quickly can I get this done so I can work on another thing?

It would take me a lot less time to take a photograph and run it through a photo editing tool to detect edges and emulate a "hand drawn" look. That is, if that's what I was going for.

That's cool, I suppose.

Production

I have touched on this topic before, and I'll likely write about it more clearly at a later time, but I do want to point out a difference between creating and producing.

The former is about process, the latter is about results.

When creating, you're not afraid of constraints. Sometimes, the more constraints you have, the more creative you become (i.e., drawing with a single pencil instead of an array of colors and paints).

When producing, you likely would rather have all constraints removed.

In my writing, it's easy to create a dichotomy between the two—without recognizing that most of the time, both are necessary. It's easy to write in ideals. It's harder to negotiate the reality.

In this post, I'm mostly focusing on the creative process as it relates outside of my work obligations. If I have free time, I want to spend it doing things I enjoy (even if they're frustrating sometimes).

For me, a shortcut is antithetical to the creative process.

I do want to point out—it's not lost on me that having time for failure is in itself a luxury.

That may be why some people may be willing to trade money for something that promises to give them more time to do the things that matter.

This is paradoxical.

You might produce more, but you're still not going to get more time. You'll just have more stuff.

I'm not buying the siren's song of tech companies pretending that automation tools will give me more time for things that matter.

Tech companies are not concerned with my time.

They are instead concerned with my attention. With my money. And my labor.

They don't want to make a tool that makes me more creative. They do want to make a tool that makes me more productive. But they most definitely don't want to give me back my time.

The Friends We Made Along The Way

There's a fairly tired TV trope where the protagonists spend a great deal of time and effort to find a thing, only to find out that said thing is either worthless or missing altogether.

The idea is that the treasure is in the journey (and the friends made along the way)!

While this sounds truthy, it is a rather dull simplification.

It is true that I never ended up with my cool Sci-Finder app (it even had a name). It's true that I learned a ton trying to achieve it. It's true that it was challenging. And it's true that I made a lot of friends along the way.

It's true that I wouldn't trade the journey for yet another semi-interesting and likely forgotten app.

But failure is valuable.

Be wary of anyone that tries to sell you something that promises to give you more time.

This post is not about eschewing automation or tools that make your life a little easier.

It's only asking you to value your constraints, flex your creativity, and embrace the gift of failure.