The Singular sins and blessings of Creators 📜

Intended audience: You may belong to the intended audience of these writings, if ... 👷‍♀️:

• you are a writer 🖋
• you are an artist 🖌
• you are an author (yes I consider computer programmers authors as well) 📰
• you are interested in creating something new 🆕
• you want to listen to my philosophical rants 📜

In general the target audience of this article is something one may refer to as a content creator.

Content Creators may create something new, original, never seen before 📝. They have sources or inspirations for their work perhaps, but their creativity and specific path of interest comes from nothing but the randomness of space 🌌.

Or is there something else to it? Will-power 👽?

When you do not create things, you become defined by your tastes rather than your ability. Your tastes only narrow & exclude people. So create. ~ Why the lucky stiff.

Sins a creator can commit 🧟‍♂️

It is important for a creator to remember he can make mistakes.

Listen to criticism, do not be afraid to apply meta-criticism back at them though if you think they are wrong.

• Pride 🏅. If a creator has too much pride in his creation he may not see flaws. Remember: Always try to be your own best critic. Then carefully listen to others points 👂. If someone talks badly about your creation do not get too angry 💢.
• Greed 📈. We all love to see the number of viewers and subscribers go up 📈. Remember your interest is what got this whole thing started from 0. Always consider consumption statistics as a byproduct of the source of the original creation. 📊
• Laziness 🌚. Resting on your laurels might be tempting. Push forward and discover the exploration drive that got your creation going. You may even discover or invent new fields of interest or expansion 🔭.
• Mind-freeze 🥶. Somehow we may get stuck in a way of thinking. Moving stubbornly in one direction may be counterproductive. "Move fast break things 🚅" might be the antidote of this mind-freeze syndrome. Always remember your creation is changeable and can be reconfigured as long as you understand its inner workings. Experiment.
• Over-communication 📢. While you create, you may release some early alpha stage of your creation before it is finished. Think before release! Do not spend too much time on communication channels talking about the finished product before it even exists. Some polish may help. Take your time to create something great 🕰 !

Plot twist 🔀:

Some of these "sins" are needed to get started with creation in the first place. Some of the sins listed even might be contradictory. I will give some examples, maybe you can come up with points for the others:

• If you have no feeling of self worth or some pride, why create in the first place? Are you a nihilist random walker destined to end up at some random destination🥉? Maybe. You found the energy to start this anyways 🌩.
• Measured laziness is needed sometimes. Rest as long as it takes, no stressful burn-out please 🥵

Do Neural networks make writers and artists obsolete? 🧠

Neural Networks have a parameter called "Temperature" that attempts to replicate that randomness that seems natural to text written by humans (or any other kind of human creation 🖌 ). If no randomness or temperature is added to the neural networks output its writings seems to become stale, repetitive and therefor less useful.

I can highly recommend Stephen Wolframs article on the matter. Especially when he talks about temperature 🌡

As large language models scale 📈 we start to bare witness to the limits of the current specific neural network technology. The most important parameters that characterize model performance are training dataset size and the number of network parameters 📀.

Performance means: The certainty with which a large language model predicts the next word. This seems to approach a limit that can not be crossed without investing exponentially more compute. The entropy of natural language and text seems to hint at something these networks can not recreate without having content remotely like it in the hidden dark and dusty corners of their training-data-set first 🕸 .

Bye

Thank you for reading my Article or rant ~ Artiosis.

- I plan to upload an emoji-free verstion of the article in the future ;)