Love Your Work

The novels that programmers generate are all over the board. One “novel” was just Moby Dick, written backwards. Another “novel” was called Paradissssse Lossssst. It was a reproduction of John Milton’s epic poem, but with each “s” in the poem replaced with a varying number of other s’s.

But, some programmers take the task a little more seriously. They train AI models and see what they come up with. One such model is called GPT-2. GPT-2 was once considered too dangerous to release to the public, because you could supposedly generate subversive content en-masse, and do some pretty nefarious things. Kind of like [Russia did with a farm of human-generated content around the 2016 election].

And what is this advanced AI model able to generate? So far, nothing impressive. Programmer and author of [aiweirdness.com] Janelle Shane tweeted, “Struggling with crafting the first sentence of your novel? Be comforted by the fact that AI is struggling even more.”

The sentence this AI model generated for Janelle: “I was playing with my dog, Mark the brown Labrador, and I had forgotten that I was also playing with a dead man.” Not exactly Tolstoy.

The follow up to GPT-2 is now out, so we’ll see this year what kind of novel GPT-3 can generate, but if Janelle Shane’s experiments so far are any indication, humans will still have the edge. She asked GPT-3 how many eyes a horse had. It kept telling her: [four].

According to Kai-Fu Lee, author of AI Superpowers, forty- to fifty-percent of jobs will be replaced by AI and automation within the next couple of decades. But humans won’t be replaced across the board. It’s the creativity- and strategy-based jobs that will be the most secure.

If your job is an “optimization-based” job, you might want to start reinventing yourself. If your primary work is maximizing a tax refund, calculating an insurance premium, or even diagnosing an illness, your job involves so-called “narrow tasks.” These tasks are already being automated, or soon will be automated.

You could type out 50,000 nonsense words in about a day. A computer can generate 50,000 words faster than you can blink. But, you could write a novel in a month. A computer can’t write a novel at all.

Which means your edge as a human is not in typing the words faster. Your edge as a human is in thinking the thoughts behind the words.

This doesn’t just apply to writing novels. If you’re an entrepreneur building a world-changing startup or a social worker helping a family navigate taking care of a sick loved-one, your creativity matters. No AI will be able to do what you do for a very long time – if ever.

So when a computer can do in the blink of an eye something that would take us all day, and when our creativity is the one thing keeping us relevant, that has powerful implications on how we get things done.

Remember from episode 226 when we learned about [Frederick Taylor]? How he stood next to a worker with a stopwatch and timed every action and broke down all of those actions into a series of steps? He optimized time as a “production unit.”

But creativity doesn’t work like stacking bricks or moving chunks of iron. Remember there are three big realities about creativity that make it incompatible with the “time management” paradigm:

1. Great ideas come in an instant
2. One idea can be infinitely more valuable than another idea
3. You can’t connect inputs directly to outputs

In a world where creativity not only matters, it’s arguably the only thing that matters, the ways that time management is incompatible with creativity are big problems.

They’re especially big problems because the more you’re watching the clock – the more you’re a [“clock-time” person], like we talked about on episode 235, the less creative you’re going to be.

So the things that used to make us more productive, now make us less productive. We can’t try to do more things in less time. We can’t multitask. We can’t skip out on sleep or otherwise neglect our health.

If you want to kill creativity: Get five hours of sleep a night, fight traffic for two hours a day, and start each day with a piping hot thermos of a psychoactive drug. This is the unfortunate and inescapable reality of most Americans today.

Will an unassisted AI be winning the Nobel Prize in literature in the next ten years? Some might think so. I’m no AI expert, but I’m skeptical. Remember from Episode 237 that [the birthday problem] shows us how hard it is for us humans to understand how complex some things are.

GPT-3 is one-hundred times more powerful than GPT-2. But is it one-hundred times better at writing a novel? We’ll see – I doubt it.

Does that make AI and other technologies useless in creative work? Far from it. We can use technology not only to lift us out of drudgery, but to assist us in being creative.

Here’s just some of the ways I use technology to be more creative:

• I live in a cheaper country, where I can have more flexibility to do work with unpredictable success ([Extremistan] like we talked about in the previous episode).
• When I moved to South America, I mourned the loss of easy access to paper books. But now, five years later, I have many thousands of highlights of the most important ideas I’ve come across in my reading. This is because I’ve been forced to read almost everything on Kindle. I can quickly and easily search through those highlights. This makes writing new books much easier than it would be otherwise.
• I’m able to live in South America because of cheap air travel, access to massive amounts of knowledge through the internet, and global publishing power, communication, and electronic banking. Not to mention easy Spanish translation in the palm of my hand.
• Aside from those Kindle highlights, I can store, organize, and quickly retrieve relevant information I’ve previously consumed or taken notes on. I can quickly reference old ideas and connect them to make new ideas.
• I’m able to test out my ideas and get instant feedback on what’s working or not through Twitter, and email, website, and podcast stats.
• Amazon’s algorithms help relevant readers find my books, which earns me money so I can write more.

There are starting to be some glimmers of AI assisting us in creativity in some more direct ways. A new service called [Sudowrite] won’t write a novel for you, but it uses GPT-3 to suggest characters or plot twists for your novel.

If you combine advances in AI models with the trends there are in studying the structure of stories, it’s not hard to see a future where AI plays a big role in assisting writers in coming up with stories.

But for now, don’t expect technology to be creative for you. Instead, use technology to help you be more creative.

New times call for new measures. When we’re trying to define what it means to be more productive, we can’t apply thinking from the industrial age when we’re in the midst of the creative age.

Image: [Traverse Beams, by Patrick Henry Bruce]

After nearly a decade of work, Mind Management, Not Time Management debuts October 27th! This book will show you how to manage your mental energy to be productive when creativity matters. Pre-order it today!

Start off each week with a dose of inspiration to help you make it as a creative. Sign up at: kadavy.net/mondays

David Kadavy is author of Mind Management, Not Time Management, The Heart to Start and Design for Hackers. Through the Love Your Work podcast, his Love Mondays newsletter, and self-publishing coaching David helps you make it as a creative.

Follow David on:

• Twitter
• Instagram
• Facebook
• YouTube

• Apple Podcasts
• Overcast
• Spotify
• Stitcher
• YouTube
• RSS
• Email