If I could tell my younger self a few rules, they would boost my career.

1. Treat design like medicine; don't get attached.

If it doesn't help, you won't be angry, but you'll try to improve it. Designers blame others if they don't like the design, but the rule is the same: we solve users' problems. You're not your design, and neither are they. Be humble with your work because your assumptions will often be wrong and users will behave differently.

2. Consider your design flawed.

Disagree with yourself, then defend your ideas. Most designers forget to dig deeper into a pattern, screen, button, or copywriting. If someone asked, "Have you considered alternatives? How does this design stack up? Here's a functional UX checklist to help you make design decisions.

3. Codeable solutions.

If your design requires more developer time, consider whether it's worth spending more money to code something with a small UX impact. Overthinking problems and designing abstract patterns is easy. Sometimes you see something on dribbble or bechance and try to recreate it, but it's not worth it. Here's my article on it.

4. Communication changes careers

Designers often talk with users, clients, companies, developers, and other designers. How you talk and present yourself can land you a job. Like driving or swimming, practice it. Success requires being outgoing and friendly. If I hadn't said "hello" to a few people, I wouldn't be where I am now.

5. Ignorance of the law is not an excuse.

Copyright, taxation How often have you used an icon without checking its license? If you use someone else's work in your project, the owner can cause you a lot of problems — paying a lot of money isn't worth it. Spend a few hours reading about copyrights, client agreements, and taxes.

6. Always test your design

If nobody has seen or used my design, it's not finished. Ask friends about prototypes. Testing reveals how wrong your assumptions were. Steve Krug, one of the authorities on this topic will tell you more about how to do testing.

7. Run workshops

A UX designer's job involves talking to people and figuring out what they need, which is difficult because they usually don't know. Organizing teamwork sessions is a powerful skill, but you must also be a good listener. Your job is to help a quiet, introverted developer express his solution and control the group. AJ Smart has more on workshops here.

This post is for you if you can't read or study for 5 minutes.

If you clicked this post, you may be experiencing problems focusing on tasks. A few minutes of reading may tire you. Easily distracted? Using social media and video games for hours without being sidetracked may impair your dopamine system.

When we achieve a goal, the brain secretes dopamine. It might be as simple as drinking water or as crucial as college admission. Situations vary. Various events require different amounts.

Dopamine is released when we start learning but declines over time. Social media algorithms provide new material continually, making us happy. Social media use slows down the system. We can't continue without an award. We return to social media and dopamine rewards.

Mice were given a button that released dopamine into their brains to study the hormone. The mice lost their hunger, thirst, and libido and kept pressing the button. Think this is like someone who spends all day gaming or on Instagram?

When we cause our brain to release so much dopamine, the brain tries to balance it in 2 ways:

1- Decreases dopamine production

2- Dopamine cannot reach its target.

Too many quick joys aren't enough. We'll want more joys. Drugs and alcohol are similar. Initially, a beer will get you drunk. After a while, 3-4 beers will get you drunk.

Social media is continually changing. Updates to these platforms keep us interested. When social media conditions us, we can't read a book.

Same here. I used to complete a book in a day and work longer without distraction. Now I'm addicted to Instagram. Daily, I spend 2 hours on social media. This must change. My life needs improvement. So I started the 50-day challenge.

I've compiled three dopamine-related methods.

Recommendations:

1. Day-long dopamine detox

First, take a day off from all your favorite things. Social media, gaming, music, junk food, fast food, smoking, alcohol, friends. Take a break.

Hanging out with friends or listening to music may seem pointless. Our minds are polluted. One day away from our pleasures can refresh us.

2. One-week dopamine detox by selecting

Choose one or more things to avoid. Social media, gaming, music, junk food, fast food, smoking, alcohol, friends. Try a week without Instagram or Twitter. I use this occasionally.

1. One week all together

One solid detox week. It's the hardest program. First or second options are best for dopamine detox. Time will help you.

You can walk, read, or pray during a dopamine detox. Many options exist. If you want to succeed, you must avoid instant gratification. Success after hard work is priceless.

The Lex Fridman podcast host has interviewed Elon Musk.

Lex is a minimalist YouTuber. His videos are sloppy. Suits are his trademark.

In a video, he shares a typical day. I've smashed my 6-month goals using its ideas.

Here's his schedule.

Morning Mantra

Not woo-woo. Lex's mantra reflects his practicality.

Four parts.

Rulebook

"I remember the game's rules," he says.

Among them:

• Sleeping 6–8 hours nightly

• 1–3 times a day, he checks social media.

• Every day, despite pain, he exercises. "I exercise uninjured body parts."

Visualize

He imagines his day. "Like Sims..."

He says three things he's grateful for and contemplates death.

"Today may be my last"

Objectives

Then he visualizes his goals. He starts big. Five-year goals.

Short-term goals follow. Lex says they're year-end goals.

Near but out of reach.

Principles

He lists his principles. Assertions. His goals.

He acknowledges his cliche beliefs. Compassion, empathy, and strength are key.

Here's my mantra routine:

Four-Hour Deep Work

Lex begins a four-hour deep work session after his mantra routine. Today's toughest.

AI is Lex's specialty. His video doesn't explain what he does.

Clearly, he works hard.

Before starting, he has water, coffee, and a bathroom break.

"During deep work sessions, I minimize breaks."

He's distraction-free. Phoneless. Silence. Nothing. Any loose ideas are typed into a Google doc for later. He wants to work.

"Just get the job done. Don’t think about it too much and feel good once it’s complete." — Lex Fridman

30-Minute Social Media & Music

After his first deep work session, Lex rewards himself.

10 minutes on social media, 20 on music. Upload content and respond to comments in 10 minutes. 20 minutes for guitar or piano.

"In the real world, I’m currently single, but in the music world, I’m in an open relationship with this beautiful guitar. Open relationship because sometimes I cheat on her with the acoustic." — Lex Fridman

Two-hour exercise

Then exercise for two hours.

Daily runs six miles. Then he chooses how far to go. Run time is an hour.

He does bodyweight exercises. Every minute for 15 minutes, do five pull-ups and ten push-ups. It's David Goggins-inspired. He aims for an hour a day.

He's hungry. Before running, he takes a salt pill for electrolytes.

He'll then take a one-minute cold shower while listening to cheesy songs. Afterward, he might eat.

Four-Hour Deep Work

Lex's second work session.

He works 8 hours a day.

Again, zero distractions.

Eating

The video's meal doesn't look appetizing, but it's healthy.

It's ground beef with vegetables. Cauliflower is his "ground-floor" veggie. "Carrots are my go-to party food."

Lex's keto diet includes 1800–2000 calories.

He drinks a "nutrient-packed" Atheltic Greens shake and takes tablets. It's:

• One daily tablet of sodium.

• Magnesium glycinate tablets stopped his keto headaches.

• Potassium — "For electrolytes"

• Fish oil: healthy joints

“So much of nutrition science is barely a science… I like to listen to my own body and do a one-person, one-subject scientific experiment to feel good.” — Lex Fridman

Four-hour shallow session

This work isn't as mentally taxing.

Lex planned to:

• Finish last session's deep work (about an hour)

• Adobe Premiere podcasting (about two hours).

• Email-check (about an hour). Three times a day max. First, check for emergencies.

If he's sick, he may watch Netflix or YouTube documentaries or visit friends.

“The possibilities of chaos are wide open, so I can do whatever the hell I want.” — Lex Fridman

Two-hour evening reading

Nonstop work.

Lex ends the day reading academic papers for an hour. "Today I'm skimming two machine learning and neuroscience papers"

This helps him "think beyond the paper."

He reads for an hour.

“When I have a lot of energy, I just chill on the bed and read… When I’m feeling tired, I jump to the desk…” — Lex Fridman

Takeaways

Lex's day-in-the-life video is inspiring.

He has positive energy and works hard every day.

Schedule:

• Mantra Routine includes rules, visualizing, goals, and principles.

• Deep Work Session #1: Four hours of focus.

• 10 minutes social media, 20 minutes guitar or piano. "Music brings me joy"

• Six-mile run, then bodyweight workout. Two hours total.

• Deep Work #2: Four hours with no distractions. Google Docs stores random thoughts.

• Lex supplements his keto diet.

• This four-hour session is "open to chaos."

• Evening reading: academic papers followed by fiction.

"I value some things in life. Work is one. The other is loving others. With those two things, life is great." — Lex Fridman

Don't sleep with your own sister.

70% of lottery winners go broke within five years. You've heard the last one. People who got rich quickly without setbacks and hard work often lose it all. My father said, "Easy money is easily lost," and a wealthy friend who owns a family office said, "The first generation makes it, the second generation spends it, and the third generation blows it."

This is evident. Corrupt politicians in developing countries live lavishly, buying their third wives' fifth Hermès bag and celebrating New Year's at The Brando Resort. A successful businessperson from humble beginnings is more conservative with money. More so if they're atom-based, not bit-based. They value money.

Crypto can "feel" easy. I have nothing against capital market investing. The global financial system is shady, but that's another topic. The problem started when those who took advantage of easy money started affecting other businesses. VCs did minimal due diligence on FTX because they needed deal flow and returns for their LPs. Lenders did minimum diligence and underwrote ludicrous loans to 3AC because they needed revenue.

Alameda (hence FTX) and 3AC made "easy money" Genesis and DCG aren't. Their businesses are more conventional, but they underestimated how "easy money" can hurt them.

Genesis has been the victim of easy money hubris and insolvency, losing $1 billion+ to 3AC and $200M to FTX. We discuss the implications for the broader crypto market.

Here are the quick takeaways:

• Genesis is one of the largest and most notable crypto lenders and prime brokerage firms.

• DCG and Genesis have done related party transactions, which can be done right but is a bad practice.

• Genesis owes DCG $1.5 billion+.

• If DCG unwinds Grayscale's GBTC, $9-10 billion in BTC will hit the market.

• DCG will survive Genesis.

What happened?

Let's recap the FTX shenanigan from two weeks ago. Shenanigans! Delphi's tweet sums up the craziness. Genesis has $175M in FTX.

Cred's timeline: I hate bad crisis management. Yes, admitting their balance sheet hole right away might've sparked more panic, and there's no easy way to convey your trouble, but no one ever learns.

By November 23, rumors circulated online that the problem could affect Genesis' parent company, DCG. To address this, Barry Silbert, Founder, and CEO of DCG released a statement to shareholders.

• A few things are confirmed thanks to this statement.

• DCG owes $1.5 billion+ to Genesis.

• $500M is due in 6 months, and the rest is due in 2032 (yes, that’s not a typo).

• Unless Barry raises new cash, his last-ditch efforts to repay the money will likely push the crypto market lower.

• Half a year of GBTC fees is approximately $100M.

• They can pay $500M with GBTC.

• With profits, sell another port.

Genesis has hired a restructuring adviser, indicating it is in trouble.

Rehypothecation

Every crypto problem in the past year seems to be rehypothecation between related parties, excessive leverage, hubris, and the removal of the money printer. The Bankless guys provided a chart showing 2021 crypto yield.

In June 2022, @DataFinnovation published a great investigation about 3AC and DCG. Here's a summary.

• 3AC borrowed BTC from Genesis and pledged it to create Grayscale's GBTC shares.

• 3AC uses GBTC to borrow more money from Genesis.

• This lets 3AC leverage their capital.

• 3AC's strategy made sense because GBTC had a premium, creating "free money."

• GBTC's discount and LUNA's implosion caused problems.

• 3AC lost its loan money in LUNA.

• Margin called on 3ACs' GBTC collateral.

• DCG bought GBTC to avoid a systemic collapse and a larger discount.

• Genesis lost too much money because 3AC can't pay back its loan. DCG "saved" Genesis, but the FTX collapse hurt Genesis further, forcing DCG and Genesis to seek external funding.

bruh…

Learning Experience

Co-borrowing. Unnecessary rehypothecation. Extra space. Governance disaster. Greed, hubris. Crypto has repeatedly shown it can recreate traditional financial system disasters quickly. Working in crypto is one of the best ways to learn crazy financial tricks people will do for a quick buck much faster than if you dabble in traditional finance.

Moving Forward

I think the crypto industry needs to consider its future. This is especially true for professionals. I'm not trying to scare you. In 2018 and 2020, I had doubts. No doubts now. Detailing the crypto industry's potential outcomes helped me gain certainty and confidence in its future. This includes VCs' benefits and talking points during the bull market, as well as what would happen if government regulations became hostile, etc. Even if that happens, I'm certain. This is permanent. I may write a post about that soon.

Sincerely,

M.

No other calendar is simpler, smaller, and reusable year after year. It works and is used here.

Most of us discard and replace our calendars annually. Each month, we move our calendar ahead another page, thus if we need to know which day of the week corresponds to a given day/month combination, we have to calculate it or flip forward/backward to the corresponding month. Questions like:

• What day does this year's American Thanksgiving fall on?

• Which months contain a Friday the thirteenth?

• When is July 4th? What day of the week?

• Alternatively, what day of the week is Christmas?

They're hard to figure out until you switch to the right month or look up all the months.

However, mathematically, the answers to these questions or any question that requires matching the day of the week with the day/month combination in a year are predictable, basic, and easy to work out. If you use this one-page calendar instead of a 12-month calendar, it lasts the whole year and is easy to alter for future years. Let me explain.

The 2023 one-page calendar is above. The days of the month are on the lower left, which works for all months if you know that:

• There are 31 days in January, March, May, July, August, October, and December.

• All of the months of April, June, September, and November have 30 days.

• And depending on the year, February has either 28 days (in non-leap years) or 29 days (in leap years).

If you know this, this calendar makes it easy to match the day/month of the year to the weekday.

Here are some instances. American Thanksgiving is always on the fourth Thursday of November. You'll always know the month and day of the week, but the date—the day in November—changes each year.

On any other calendar, you'd have to flip to November to see when the fourth Thursday is. This one-page calendar only requires:

• pick the month of November in the top-right corner to begin.

• drag your finger down until Thursday appears,

• then turn left and follow the monthly calendar until you reach the fourth Thursday.

It's obvious: 2023 is the 23rd American Thanksgiving. For every month and day-of-the-week combination, start at the month, drag your finger down to the desired day, and then move to the left to see which dates match.

What if you knew the day of the week and the date of the month, but not the month(s)?

A different method using the same one-page calendar gives the answer. Which months have Friday the 13th this year? Just:

• begin on the 13th of the month, the day you know you desire,

• then swipe right with your finger till Friday appears.

• and then work your way up until you can determine which months the specific Friday the 13th falls under.

One Friday the 13th occurred in January 2023, and another will occur in October.

The most typical reason to consult a calendar is when you know the month/day combination but not the day of the week.

Compared to single-month calendars, the one-page calendar excels here. Take July 4th, for instance. Find the weekday here:

• beginning on the left on the fourth of the month, as you are aware,

• also begin with July, the month of the year you are most familiar with, at the upper right,

• you should move your two fingers in the opposite directions till they meet: on a Tuesday in 2023.

That's how you find your selected day/month combination's weekday.

Another example: Christmas. Christmas Day is always December 25th, however unless your conventional calendar is open to December of your particular year, a question like "what day of the week is Christmas?" difficult to answer.

Unlike the one-page calendar!

Remember the left-hand day of the month. Top-right, you see the month. Put two fingers, one from each hand, on the date (25th) and the month (December). Slide the day hand to the right and the month hand downwards until they touch.

They meet on Monday—December 25, 2023.

For 2023, that's fine, but what happens in 2024? Even worse, what if we want to know the day-of-the-week/day/month combo many years from now?

I think the one-page calendar shines here.

Except for the blue months in the upper-right corner of the one-page calendar, everything is the same year after year. The months also change in a consistent fashion.

Each non-leap year has 365 days—one more than a full 52 weeks (which is 364). Since January 1, 2023 began on a Sunday and 2023 has 365 days, we immediately know that December 31, 2023 will conclude on a Sunday (which you can confirm using the one-page calendar) and that January 1, 2024 will begin on a Monday. Then, reorder the months for 2024, taking in mind that February will have 29 days in a leap year.

Please note the differences between 2023 and 2024 month placement. In 2023:

• October and January began on the same day of the week.

• On the following Monday of the week, May began.

• August started on the next day,

• then the next weekday marked the start of February, March, and November, respectively.

• Unlike June, which starts the following weekday,

• While September and December start on the following day of the week,

• Lastly, April and July start one extra day later.

Since 2024 is a leap year, February has 29 days, disrupting the rhythm. Month placements change to:

• The first day of the week in January, April, and July is the same.

• October will begin the following day.

• Possibly starting the next weekday,

• February and August start on the next weekday,

• beginning on the following day of the week between March and November,

• beginning the following weekday in June,

• and commencing one more day of the week after that, September and December.

Due to the 366-day leap year, 2025 will start two days later than 2024 on January 1st.

Now, looking at the 2025 calendar, you can see that the 2023 pattern of which months start on which days is repeated! The sole variation is a shift of three days-of-the-week ahead because 2023 had one more day (365) than 52 full weeks (364), and 2024 had two more days (366). Again,

• On Wednesday this time, January and October begin on the same day of the week.

• Although May begins on Thursday,

• August begins this Friday.

• March, November, and February all begin on a Saturday.

• Beginning on a Sunday in June

• Beginning on Monday are September and December,

• and on Tuesday, April and July begin.

In 2026 and 2027, the year will commence on a Thursday and a Friday, respectively.

We must return to our leap year monthly arrangement in 2028. Yes, January 1, 2028 begins on a Saturday, but February, which begins on a Tuesday three days before January, will have 29 days. Thus:

• Start dates for January, April, and July are all Saturdays.

• Given that October began on Sunday,

• Although May starts on a Monday,

• beginning on a Tuesday in February and August,

• Beginning on a Wednesday in March and November,

• Beginning on Thursday, June

• and Friday marks the start of September and December.

This is great because there are only 14 calendar configurations: one for each of the seven non-leap years where January 1st begins on each of the seven days of the week, and one for each of the seven leap years where it begins on each day of the week.

The 2023 calendar will function in 2034, 2045, 2051, 2062, 2073, 2079, 2090, 2102, 2113, and 2119. Except when passing over a non-leap year that ends in 00, like 2100, the repeat time always extends to 12 years or shortens to an extra 6 years.

• The pattern is repeated in 2025's calendar in 2031, 2042, 2053, 2059, 2070, 2081, 2087, 2098, 2110, and 2121.

• The extra 6-year repeat at the end of the century on the calendar for 2026 will occur in the years 2037, 2043, 2054, 2065, 2071, 2082, 2093, 2099, 2105, and 2122.

• The 2027s calendar repeats in 2038, 2049, 2055, 2066, 2077, 2083, 2094, 2100, 2106, and 2117, almost exactly matching the 2026s pattern.

For leap years, the recurrence pattern is every 28 years when not passing a non-leap year ending in 00, or 12 or 40 years when we do. 2024's calendar repeats in 2052, 2080, 2120, 2148, 2176, and 2216; 2028's in 2056, 2084, 2124, 2152, 2180, and 2220.

Knowing January 1st and whether it's a leap year lets you construct a one-page calendar for any year. Try it—you might find it easier than any other alternative!

Deep learning-based generative art is being researched. As usual, self-learning is better. Some models, like OpenAI's DALL-E 2, require registration and can only be used online, but others can be used locally, which is usually more enjoyable for curious users. I'll demonstrate the Stable Diffusion model's operation on a standard PC.

Let’s get started.

What It Does

Stable Diffusion uses numerous components:

• A generative model trained to produce images is called a diffusion model. The model is incrementally improving the starting data, which is only random noise. The model has an image, and while it is being trained, the reversed process is being used to add noise to the image. Being able to reverse this procedure and create images from noise is where the true magic is (more details and samples can be found in the paper).

• An internal compressed representation of a latent diffusion model, which may be altered to produce the desired images, is used (more details can be found in the paper). The capacity to fine-tune the generation process is essential because producing pictures at random is not very attractive (as we can see, for instance, in Generative Adversarial Networks).

• A neural network model called CLIP (Contrastive Language-Image Pre-training) is used to translate natural language prompts into vector representations. This model, which was trained on 400,000,000 image-text pairs, enables the transformation of a text prompt into a latent space for the diffusion model in the scenario of stable diffusion (more details in that paper).

This figure shows all data flow:

The weights file size for Stable Diffusion model v1 is 4 GB and v2 is 5 GB, making the model quite huge. The v1 model was trained on 256x256 and 512x512 LAION-5B pictures on a 4,000 GPU cluster using over 150.000 NVIDIA A100 GPU hours. The open-source pre-trained model is helpful for us. And we will.

Install

Before utilizing the Python sources for Stable Diffusion v1 on GitHub, we must install Miniconda (assuming Git and Python are already installed):

wget https://repo.anaconda.com/miniconda/Miniconda3-py39_4.12.0-Linux-x86_64.sh
chmod +x Miniconda3-py39_4.12.0-Linux-x86_64.sh
./Miniconda3-py39_4.12.0-Linux-x86_64.sh
conda update -n base -c defaults conda

Install the source and prepare the environment:

git clone https://github.com/CompVis/stable-diffusion
cd stable-diffusion
conda env create -f environment.yaml
conda activate ldm
pip3 install transformers --upgrade

Download the pre-trained model weights next. HiggingFace has the newest checkpoint sd-v14.ckpt (a download is free but registration is required). Put the file in the project folder and have fun:

python3 scripts/txt2img.py --prompt "hello world" --plms --ckpt sd-v1-4.ckpt --skip_grid --n_samples 1

Almost. The installation is complete for happy users of current GPUs with 12 GB or more VRAM. RuntimeError: CUDA out of memory will occur otherwise. Two solutions exist.

Running the optimized version

Try optimizing first. After cloning the repository and enabling the environment (as previously), we can run the command:

python3 optimizedSD/optimized_txt2img.py --prompt "hello world" --ckpt sd-v1-4.ckpt --skip_grid --n_samples 1

Stable Diffusion worked on my visual card with 8 GB RAM (alas, I did not behave well enough to get NVIDIA A100 for Christmas, so 8 GB GPU is the maximum I have;).

Running Stable Diffusion without GPU

If the GPU does not have enough RAM or is not CUDA-compatible, running the code on a CPU will be 20x slower but better than nothing. This unauthorized CPU-only branch from GitHub is easiest to obtain. We may easily edit the source code to use the latest version. It's strange that a pull request for that was made six months ago and still hasn't been approved, as the changes are simple. Readers can finish in 5 minutes:

• Replace if attr.device!= torch.device(cuda) with if attr.device!= torch.device(cuda) and torch.cuda.is available at line 20 of ldm/models/diffusion/ddim.py ().

• Replace if attr.device!= torch.device(cuda) with if attr.device!= torch.device(cuda) and torch.cuda.is available in line 20 of ldm/models/diffusion/plms.py ().

• Replace device=cuda in lines 38, 55, 83, and 142 of ldm/modules/encoders/modules.py with device=cuda if torch.cuda.is available(), otherwise cpu.

• Replace model.cuda() in scripts/txt2img.py line 28 and scripts/img2img.py line 43 with if torch.cuda.is available(): model.cuda ().

Run the script again.

Testing

Test the model. Text-to-image is the first choice. Test the command line example again:

python3 scripts/txt2img.py --prompt "hello world" --plms --ckpt sd-v1-4.ckpt --skip_grid --n_samples 1

The slow generation takes 10 seconds on a GPU and 10 minutes on a CPU. Final image:

Hello world is dull and abstract. Try a brush-wielding hamster. Why? Because we can, and it's not as insane as Napoleon's cat. Another image:

Generating an image from a text prompt and another image is interesting. I made this picture in two minutes using the image editor (sorry, drawing wasn't my strong suit):

I can create an image from this drawing:

python3 scripts/img2img.py --prompt "A bird is sitting on a tree branch" --ckpt sd-v1-4.ckpt --init-img bird.png --strength 0.8

It was far better than my initial drawing:

I hope readers understand and experiment.

Stable Diffusion UI

Developers love the command line, but regular users may struggle. Stable Diffusion UI projects simplify image generation and installation. Simple usage:

• Unpack the ZIP after downloading it from https://github.com/cmdr2/stable-diffusion-ui/releases. Linux and Windows are compatible with Stable Diffusion UI (sorry for Mac users, but those machines are not well-suitable for heavy machine learning tasks anyway;).

• Start the script.

Done. The web browser UI makes configuring various Stable Diffusion features (upscaling, filtering, etc.) easy:

V2.1 of Stable Diffusion

I noticed the notification about releasing version 2.1 while writing this essay, and it was intriguing to test it. First, compare version 2 to version 1:

• alternative text encoding. The Contrastive LanguageImage Pre-training (CLIP) deep learning model, which was trained on a significant number of text-image pairs, is used in Stable Diffusion 1. The open-source CLIP implementation used in Stable Diffusion 2 is called OpenCLIP. It is difficult to determine whether there have been any technical advancements or if legal concerns were the main focus. However, because the training datasets for the two text encoders were different, the output results from V1 and V2 will differ for the identical text prompts.

• a new depth model that may be used to the output of image-to-image generation.

• a revolutionary upscaling technique that can quadruple the resolution of an image.

• Generally higher resolution Stable Diffusion 2 has the ability to produce both 512x512 and 768x768 pictures.

The Hugging Face website offers a free online demo of Stable Diffusion 2.1 for code testing. The process is the same as for version 1.4. Download a fresh version and activate the environment:

conda deactivate  
conda env remove -n ldm  # Use this if version 1 was previously installed
git clone https://github.com/Stability-AI/stablediffusion
cd stablediffusion
conda env create -f environment.yaml
conda activate ldm

Hugging Face offers a new weights ckpt file.

The Out of memory error prevented me from running this version on my 8 GB GPU. Version 2.1 fails on CPUs with the slow conv2d cpu not implemented for Half error (according to this GitHub issue, the CPU support for this algorithm and data type will not be added). The model can be modified from half to full precision (float16 instead of float32), however it doesn't make sense since v1 runs up to 10 minutes on the CPU and v2.1 should be much slower. The online demo results are visible. The same hamster painting with a brush prompt yielded this result:

It looks different from v1, but it functions and has a higher resolution.

The superresolution.py script can run the 4x Stable Diffusion upscaler locally (the x4-upscaler-ema.ckpt weights file should be in the same folder):

python3 scripts/gradio/superresolution.py configs/stable-diffusion/x4-upscaling.yaml x4-upscaler-ema.ckpt

This code allows the web browser UI to select the image to upscale:

The copy-paste strategy may explain why the upscaler needs a text prompt (and the Hugging Face code snippet does not have any text input as well). I got a GPU out of memory error again, although CUDA can be disabled like v1. However, processing an image for more than two hours is unlikely:

Stable Diffusion Limitations

When we use the model, it's fun to see what it can and can't do. Generative models produce abstract visuals but not photorealistic ones. This fundamentally limits The generative neural network was trained on text and image pairs, but humans have a lot of background knowledge about the world. The neural network model knows nothing. If someone asks me to draw a Chinese text, I can draw something that looks like Chinese but is actually gibberish because I never learnt it. Generative AI does too! Humans can learn new languages, but the Stable Diffusion AI model includes only language and image decoder brain components. For instance, the Stable Diffusion model will pull NO WAR banner-bearers like this:

V1:

V2.1:

The shot shows text, although the model never learned to read or write. The model's string tokenizer automatically converts letters to lowercase before generating the image, so typing NO WAR banner or no war banner is the same.

I can also ask the model to draw a gorgeous woman:

V1:

V2.1:

The first image is gorgeous but physically incorrect. A second one is better, although it has an Uncanny valley feel. BTW, v2 has a lifehack to add a negative prompt and define what we don't want on the image. Readers might try adding horrible anatomy to the gorgeous woman request.

If we ask for a cartoon attractive woman, the results are nice, but accuracy doesn't matter:

V1:

V2.1:

Another example: I ordered a model to sketch a mouse, which looks beautiful but has too many legs, ears, and fingers:

V1:

V2.1: improved but not perfect.

V1 produces a fun cartoon flying mouse if I want something more abstract:

I tried multiple times with V2.1 but only received this:

The image is OK, but the first version is closer to the request.

Stable Diffusion struggles to draw letters, fingers, etc. However, abstract images yield interesting outcomes. A rural landscape with a modern metropolis in the background turned out well:

V1:

V2.1:

Generative models help make paintings too (at least, abstract ones). I searched Google Image Search for modern art painting to see works by real artists, and this was the first image:

I typed "abstract oil painting of people dancing" and got this:

V1:

V2.1:

It's a different style, but I don't think the AI-generated graphics are worse than the human-drawn ones.

The AI model cannot think like humans. It thinks nothing. A stable diffusion model is a billion-parameter matrix trained on millions of text-image pairs. I input "robot is creating a picture with a pen" to create an image for this post. Humans understand requests immediately. I tried Stable Diffusion multiple times and got this:

This great artwork has a pen, robot, and sketch, however it was not asked. Maybe it was because the tokenizer deleted is and a words from a statement, but I tried other requests such robot painting picture with pen without success. It's harder to prompt a model than a person.

I hope Stable Diffusion's general effects are evident. Despite its limitations, it can produce beautiful photographs in some settings. Readers who want to use Stable Diffusion results should be warned. Source code examination demonstrates that Stable Diffusion images feature a concealed watermark (text StableDiffusionV1 and SDV2) encoded using the invisible-watermark Python package. It's not a secret, because the official Stable Diffusion repository's test watermark.py file contains a decoding snippet. The put watermark line in the txt2img.py source code can be removed if desired. I didn't discover this watermark on photographs made by the online Hugging Face demo. Maybe I did something incorrectly (but maybe they are just not using the txt2img script on their backend at all).

Conclusion

The Stable Diffusion model was fascinating. As I mentioned before, trying something yourself is always better than taking someone else's word, so I encourage readers to do the same (including this article as well;).

Is Generative AI a game-changer? My humble experience tells me:

• I think that place has a lot of potential. For designers and artists, generative AI can be a truly useful and innovative tool. Unfortunately, it can also pose a threat to some of them since if users can enter a text field to obtain a picture or a website logo in a matter of clicks, why would they pay more to a different party? Is it possible right now? unquestionably not yet. Images still have a very poor quality and are erroneous in minute details. And after viewing the image of the stunning woman above, models and fashion photographers may also unwind because it is highly unlikely that AI will replace them in the upcoming years.

• Today, generative AI is still in its infancy. Even 768x768 images are considered to be of a high resolution when using neural networks, which are computationally highly expensive. There isn't an AI model that can generate high-resolution photographs natively without upscaling or other methods, at least not as of the time this article was written, but it will happen eventually.

• It is still a challenge to accurately represent knowledge in neural networks (information like how many legs a cat has or the year Napoleon was born). Consequently, AI models struggle to create photorealistic photos, at least where little details are important (on the other side, when I searched Google for modern art paintings, the results are often even worse;).

• When compared to the carefully chosen images from official web pages or YouTube reviews, the average output quality of a Stable Diffusion generation process is actually less attractive because to its high degree of randomness. When using the same technique on their own, consumers will theoretically only view those images as 1% of the results.

Anyway, it's exciting to witness this area's advancement, especially because the project is open source. Google's Imagen and DALL-E 2 can also produce remarkable findings. It will be interesting to see how they progress.