Redefining what's possible
Redefining what's possible
And how it looks in practice
Hello friend.
This week, an important distinction I want you to know and progress on the story-based course putting that into practice.
Let’s get into it.
Main quest 🛡
This time, I want to talk to you about the “impossible”.
All of this comes from a question that I asked myself a few years ago.
“What does it take to do the impossible?”
And a related one.
“What is it that happens for the impossible to become possible?”
Since this is a long topic, I want to give you one of the main points.
That’s the distinction between “impossible” and “IMPOSSIBLE”.
(And is not only that one is lowercase and the other is uppercase. 😆)
The difference is in the reach.
An “impossible” can be called a personal impossible, something that not only you can’t currently but don’t think it’s possible for you.
Practical example, 5 years ago, I was an up and coming web developer. I was trying to get my first role and break into the tech industry.
If you were to tell me that I would become a technical writer and get paid for written articles in the tech space, I would have called you insane and delusional.
And one year later, that would’ve been the case still. Being a technical writer and knowing how to explain complex technical concepts in a simple and entertaining way was a total “impossible” for me.
The other version is an “IMPOSSIBLE”. This is a general one. It’s not only impossible for one person but for a whole group of people.
Humans flying was once in that category.
Reaching the surface of the moon was also there.
Running a sub-4-minute mile was another one. And it was once believed that it was impossible for a human to do it. Since there were a lot of people who tried it and failed.
Until, of course, it wasn’t.
Then came Roger Bannister in 1954 and became the first person to break that record.
And the most interesting part of that story is not that Bannister was the first person who made the “impossible” possible.
For me, the most interesting part is that the longstanding record Bannister broke and the new standard he set, only lasted 46 days. 😂
The “impossible” was not only possible but it was beaten again and again over the following years.
As of June 2022, the “4-minute barrier” has been broken by 1,755 athletes.
So there are “personal impossibles” and “collective impossibles”.
And we have a chance to contribute to both. Although we have to start with our own personal ones first.
If you remember from my personal example above, there’s one main important thing that helped. That is the capacity to absorb new knowledge and develop new skills in a short time span.
In other words, that is the capacity for “supercharged learning”.
You have your own personal impossibles.
You may or may not be aware of those right now, but they are there.
And when you set out to conquer those stormy seas or summit those high peaks.
Your capacity to adapt, develop, and learn will be the key that will open the door and help you get to the other side. 🚪
Experiments 🧪
I’ve been taking a stab at my own personal impossible this week.
That’s because I wanted to make an animated story for the online course I’m creating.
And you know what they say, “a picture is worth a thousand words”.
So, I started experimenting with Stable Diffusion to create the images for that “animated” part of the story. I wanted to make a short video in the anime style that many of you and folks on Twitter told me they preferred.
But I definitely overestimated the complexity of such a task.
I took me between 12 to 15 hours *just* prompting models to get the images I needed for the scenes.
And bear in mind that before this project, I had very little knowledge about image generators or how to use the different versions of Stable Diffusion.
I also didn’t know much about:
Aspect ratios.
Shot sizes.
Camera lens.
Artistic styles.
Lighting descriptions (so that the images look realistic)
Image generation models
Prompt weights
Embeddings
LoRAs (Low-Rank Adaptations)
And much more stuff.
Even with the newest version of Stable Diffusion, the one called “XL”, there’s still more work to be done so that the images generated don’t appear like portraits or the character doesn’t look at the camera like a model.
And one of the most frustrating parts is that even with great prompts (and negative prompts), Stable Diffusion takes your description and creates whatever the hell it wants. 🙄
Saying that I’ve been “experimenting” is really an understatement.
Power-ups ✨
This week, I have a very particular power-up for you.
It helped me a lot creating the graphics for the story. And even though I used like 5 different “free” text-to-image platforms.
There’s a pretty unknown one that helped me land at the anime style I was looking for.
The best part? It doesn’t limit the amount of models and styles you can try like Midjourney does.
It’s called Happy Accidents.
(Which is funny because finding about it was quite an accident tbh.)
I went deep into the SD rabbit hole even to find out that you can use all models and parameters from a user interface directly running in your own computer.
But I have a potato for PC, so I cannot run it on my machine.
That’s when that site came in and saved the day.
I used it in tandem with Leonardo AI to get a consistent style, a consistent character, and also enough images to train a model so that I can get the body poses and facial expressions needed for the rest of the story.
If you ever need to create images with a free generator and a generous tier, take a look at it.
From the vault 🏛️
There’s no additional resource coming your way this time.
But I think it’s enough with what I shared with you in the main quest already.
There will be another resource coming next week.
That is all for this time, my friend.
I’m going to need a great amount of rest after tackling that impossible.
Take care and have an awesome week ahead.
Cheers,
Juan.