It's kids' digital crack

TikTok is a destructive social media platform.

• The interface shortens attention spans and dopamine receptors.

• TikTok shares more data than other apps.

• Seeing an endless stream of dancing teens on my glowing box makes me feel like a Blade Runner extra.

TikTok did in one year what MTV, Hollywood, and Warner Music tried to do in 20 years. TikTok has psychotized the two-thirds of society Aldous Huxley said were hypnotizable.

Millions of people, mostly kids, are addicted to learning a new dance, lip-sync, or prank, and those who best dramatize this collective improvisation get likes, comments, and shares.

TikTok is a great app. So what?

The Commercial Magnifying Glass TikTok made me realize my generation's time was up and the teenage Zoomers were the target.

I told my 14-year-old sister, "Enjoy your time under the commercial magnifying glass."

TikTok sells your every move, gesture, and thought. Data is the new oil. If you tell someone, they'll say, "Yeah, they collect data, but who cares? I have nothing to hide."

It's a George Orwell novel's beginning. Look up Big Brother Award winners to see if TikTok won.

TikTok shares your data more than any other social media app, and where it goes is unclear. TikTok uses third-party trackers to monitor your activity after you leave the app.

Consumers can't see what data is shared or how it will be used. — Genius URL

32.5 percent of Tiktok's users are 10 to 19 and 29.5% are 20 to 29.

TikTok is the greatest digital marketing opportunity in history, and they'll use it to sell you things, track you, and control your thoughts. Any of its users will tell you, "I don't care, I just want to be famous."

TikTok manufactures mental illness

TikTok's effect on dopamine and the brain is absurd. Dopamine controls the brain's pleasure and reward centers. It's like a switch that tells your brain "this feels good, repeat."

Dr. Julie Albright, a digital culture and communication sociologist, said TikTok users are "carried away by dopamine." It's hypnotic, you'll keep watching."

TikTok constantly releases dopamine. A guy on TikTok recently said he didn't like books because they were slow and boring.

The US didn't ban Tiktok.

Biden and Trump agree on bad things. Both agree that TikTok threatens national security and children's mental health.

The Chinese Communist Party owns and operates TikTok, but that's not its only problem.

• There’s borderline child porn on TikTok

• It's unsafe for children and violated COPPA.

• It's also Chinese spyware. I'm not a Trump supporter, but I was glad he wanted TikTok regulated and disappointed when he failed.

Full-on internet censorship is rare outside of China, so banning it may be excessive. US should regulate TikTok more.

We must reject a low-quality present for a high-quality future.

TikTok vs YouTube

People got mad when I wrote about YouTube's death.

They didn't like when I said TikTok was YouTube's first real challenger.

Indeed. TikTok is the fastest-growing social network. In three years, the Chinese social media app TikTok has gained over 1 billion active users. In the first quarter of 2020, it had the most downloads of any app in a single quarter.

TikTok is the perfect social media app in many ways. It's brief and direct.

Can you believe they had a YouTube vs TikTok boxing match? We are doomed as a species.

YouTube hosts my favorite videos. That’s why I use it. That’s why you use it. New users expect more. They want something quicker, more addictive.

TikTok's impact on other social media platforms frustrates me. YouTube copied TikTok to compete.

It's all about short, addictive content.

I'll admit I'm probably wrong about TikTok. My friend says his feed is full of videos about food, cute animals, book recommendations, and hot lesbians.

Whatever.

TikTok makes us bad

TikTok is the opposite of what the Ancient Greeks believed about wisdom.

It encourages people to be fake. It's like a never-ending costume party where everyone competes.

It does not mean that Gen Z is doomed.

They could be the saviors of the world for all I know.

TikTok feels like a step towards Mike Judge's "Idiocracy," where the average person is a pleasure-seeking moron.

A worldwide energy crisis might have hampered renewable energy and clean tech investment. Nope.

BNEF's 2023 Energy Transition Investment Trends study surprised and encouraged. Global energy transition investment reached $1 trillion for the first time ($1.11t), up 31% from 2021. From 2013, the clean energy transition has come and cannot be reversed.

BNEF Head of Global Analysis Albert Cheung said our findings ended the energy crisis's influence on renewable energy deployment. Energy transition investment has reached a record as countries and corporations implement transition strategies. Clean energy investments will soon surpass fossil fuel investments.

The table below indicates the tripping point, which means the energy shift is occuring today.

BNEF calls money invested on clean technology including electric vehicles, heat pumps, hydrogen, and carbon capture energy transition investment. In 2022, electrified heat received $64b and energy storage $15.7b.

Nonetheless, $495b in renewables (up 17%) and $466b in electrified transport (up 54%) account for most of the investment. Hydrogen and carbon capture are tiny despite the fanfare. Hydrogen received the least funding in 2022 at $1.1 billion (0.1%).

China dominates investment. China spends $546 billion on energy transition, half the global amount. Second, the US total of $141 billion in 2022 was up 11% from 2021. With $180 billion, the EU is unofficially second. China invested 91% in battery technologies.

The 2022 transition tipping point is encouraging, but the BNEF research shows how far we must go to get Net Zero. Energy transition investment must average $4.55 trillion between 2023 and 2030—three times the amount spent in 2022—to reach global Net Zero. Investment must be seven times today's record to reach Net Zero by 2050.

BNEF 2023 Energy Transition Investment Trends.

As shown in the graph above, BNEF experts have been using their crystal balls to determine where that investment should go. CCS and hydrogen are still modest components of the picture. Interestingly, they see nuclear almost fading. Active transport advocates like me may have something to say about the massive $4b in electrified transport. If we focus on walkable 15-minute cities, we may need fewer electric automobiles. Though we need more electric trains and buses.

Albert Cheung of BNEF emphasizes the challenge. This week's figures promise short-term job creation and medium-term energy security, but more investment is needed to reach net zero in the long run.

I expect the BNEF Energy Transition Investment Trends report to show clean tech investment outpacing fossil fuels investment every year. Finally saying that is amazing. It's insufficient. The planet must maintain its electric (not gas) pedal. In response to the research, Christina Karapataki, VC at Breakthrough Energy Ventures, a clean tech investment firm, tweeted: Clean energy investment needs to average more than 3x this level, for the remainder of this decade, to get on track for BNEFs Net Zero Scenario. Go!

He’s not the President. Stop treating him like one.

Elon Musk tweeted from Qatar, where he was watching the World Cup Final with Jared Kushner.

Musk's subsequent Tweets were as normal, basic, and bland as anyone's from a World Cup Final: It's depressing to see the world's richest man looking at his phone during a grand ceremony. Rich guy goes to rich guy event didn't seem important.

Before Musk posted his should-I-step-down-at-Twitter poll, CNN ran a long segment asking if it was hypocritical for him to reveal his real-time location after defending his (very dumb) suspension of several journalists for (supposedly) revealing his assassination coordinates by linking to a site that tracks Musks private jet. It was hard to ignore CNN's hypocrisy: It covered Musk as Twitter CEO like President Trump. EVERY TRUMP STORY WAS BASED ON HIM SAYING X, THEN DOING Y. Trump would do something horrific, lie about it, then pretend it was fine, then condemn a political rival who did the same thing, be called hypocritical, and so on. It lasted four years. Exhausting.

It made sense because Trump was the President of the United States. The press's main purpose is to relentlessly cover and question the president.

It's strange to say this out. Twitter isn't America. Elon Musk isn't a president. He maintains a money-losing social media service to harass and mock people he doesn't like. Treating Musk like Trump, as if he should be held accountable like Trump, shows a startling lack of perspective. Some journalists treat Twitter like a country.

The compulsive, desperate way many journalists utilize the site suggests as much. Twitter isn't the town square, despite popular belief. It's a place for obsessives to meet and converse. Journalists say they're breaking news. Their careers depend on it. They can argue it's a public service. Nope. It's a place lonely people go to speak all day. Twitter. So do journalists, Trump, and Musk. Acting as if it has a greater purpose, as if it's impossible to break news without it, or as if the republic is in peril is ludicrous. Only 23% of Americans are on Twitter, while 25% account for 97% of Tweets. I'd think a large portion of that 25% are journalists (or attention addicts) chatting to other journalists. Their loudness makes Twitter seem more important than it is. Nope. It's another stupid website. They were there before Twitter; they will be there after Twitter. It’s just a website. We can all get off it if we want. Most of us aren’t even on it in the first place.

Musk is a website-owner. No world leader. He's not as accountable as Trump was. Musk is cable news's primary character now that Trump isn't (at least for now). Becoming a TV news anchor isn't as significant as being president. Elon Musk isn't as important as we all pretend, and Twitter isn't even close. Twitter is a dumb website, Elon Musk is a rich guy going through a midlife crisis, and cable news is lazy because its leaders thought the entire world was on Twitter and are now freaking out that their playground is being disturbed.

I’ve said before that you need to leave Twitter, now. But even if you’re still on it, we need to stop pretending it matters more than it does. It’s a site for lonely attention addicts, from the man who runs it to the journalists who can’t let go of it. It’s not a town square. It’s not a country. It’s not even a successful website. Let’s stop pretending any of it’s real. It’s not.

SpaceX's Starship.

Launched last week.

Four minutes in:

SpaceX will succeed. When it does, its massiveness will matter.

Its payload will revolutionize space economics.

Civilization will shift.

We don't yet understand how this will affect space and Earth culture. Grab it.

The Cost of Space Transportation Has Decreased Exponentially

Space launches have increased dramatically in recent years.

We mostly send items to LEO, the green area below:

SpaceX's reusable rockets can send these things to LEO. Each may launch dozens of payloads into space.

With all these launches, we're sending more than simply things to space. Volume and mass. Since the 1980s, launching a kilogram of payload to LEO has become cheaper:

One kilogram in a large rocket cost over $75,000 in the 1980s. Carrying one astronaut cost nearly $5M! Falcon Heavy's $1,500/kg price is 50 times lower. SpaceX's larger, reusable rockets are amazing.

SpaceX's Starship rocket will continue. It can carry over 100 tons to LEO, 50% more than the current Falcon heavy. Thousands of launches per year. Elon Musk predicts Falcon Heavy's $1,500/kg cost will plummet to $100 in 23 years.

In context:

People underestimate this.

2. The Benefits of Affordable Transportation

Compare Earth's transportation costs:

It's no surprise that the US and Northern Europe are the wealthiest and have the most navigable interior waterways.

So what? since sea transportation is cheaper than land. Inland waterways are even better than sea transportation since weather is less of an issue, currents can be controlled, and rivers serve two banks instead of one for coastal transportation.

In France, because population density follows river systems, rivers are valuable. Cheap transportation brought people and money to rivers, especially their confluences.

How come? Why were humans surrounding rivers?

Imagine selling meat for $10 per kilogram. Transporting one kg one kilometer costs $1. Your margin decreases $1 each kilometer. You can only ship 10 kilometers. For example, you can only trade with four cities:

If instead, your cost of transportation is half, what happens? It costs you $0.5 per km. You now have higher margins with each city you traded with. More importantly, you can reach 20-km markets.

However, 2x distance 4x surface! You can now trade with sixteen cities instead of four! Metcalfe's law states that a network's value increases with its nodes squared. Since now sixteen cities can connect to yours. Each city now has sixteen connections! They get affluent and can afford more meat.

Rivers lower travel costs, connecting many cities, which can trade more, get wealthy, and buy more.

The right network is worth at least an order of magnitude more than the left! The cheaper the transport, the more trade at a lower cost, the more income generated, the more that wealth can be reinvested in better canals, bridges, and roads, and the wealth grows even more.

Throughout history. Rome was established around cheap Mediterranean transit and preoccupied with cutting overland transportation costs with their famous roadways. Communications restricted their empire.

The Egyptians lived around the Nile, the Vikings around the North Sea, early Japan around the Seto Inland Sea, and China started canals in the 5th century BC.

Transportation costs shaped empires.Starship is lowering new-world transit expenses. What's possible?

3. Change Organizations, Change Companies, Change the World

Starship is a conveyor belt to LEO. A new world of opportunity opens up as transportation prices drop 100x in a decade.

Satellite engineers have spent decades shedding milligrams. Weight influenced every decision: pricing structure, volumes to be sent, material selections, power sources, thermal protection, guiding, navigation, and control software. Weight was everything in the mission. To pack as much science into every millimeter, NASA missions had to be miniaturized. Engineers were indoctrinated against mass.

No way.

Starship is not constrained by any space mission, robotic or crewed.

Starship obliterates the mass constraint and every last vestige of cultural baggage it has gouged into the minds of spacecraft designers. A dollar spent on mass optimization no longer buys a dollar saved on launch cost. It buys nothing. It is time to raise the scope of our ambition and think much bigger. — Casey Handmer, Starship is still not understood

A Tesla Roadster in space makes more sense.

It went beyond bad PR. It told the industry: Did you care about every microgram? No more. My rockets are big enough to send a Tesla without noticing. Industry watchers should have noticed.

Most didn’t. Artemis is a global mission to send astronauts to the Moon and build a base. Artemis uses disposable Space Launch System rockets. Instead of sending two or three dinky 10-ton crew habitats over the next decade, Starship might deliver 100x as much cargo and create a base for 1,000 astronauts in a year or two. Why not? Because Artemis remains in a pre-Starship paradigm where each kilogram costs a million dollars and we must aggressively descope our objective.

Space agencies can deliver 100x more payload to space for the same budget with 100x lower costs and 100x higher transportation volumes. How can space economy saturate this new supply?

Before Starship, NASA supplied heavy equipment for Moon base construction. After Starship, Caterpillar and Deere may space-qualify their products with little alterations. Instead than waiting decades for NASA engineers to catch up, we could send people to build a space outpost with John Deere equipment in a few years.

History is littered with the wreckage of former industrial titans that underestimated the impact of new technology and overestimated their ability to adapt: Blockbuster, Motorola, Kodak, Nokia, RIM, Xerox, Yahoo, IBM, Atari, Sears, Hitachi, Polaroid, Toshiba, HP, Palm, Sony, PanAm, Sega, Netscape, Compaq, GM… — Casey Handmer, Starship is still not understood

Everyone saw it coming, but senior management failed to realize that adaption would involve moving beyond their established business practice. Others will if they don't.

4. The Starship Possibilities

It's Starlink.

SpaceX invented affordable cargo space and grasped its implications first. How can we use all this inexpensive cargo nobody knows how to use?

Satellite communications seemed like the best way to capitalize on it. They tried. Starlink, designed by SpaceX, provides fast, dependable Internet worldwide. Beaming information down is often cheaper than cable. Already profitable.

Starlink is one use for all this cheap cargo space. Many more. The longer firms ignore the opportunity, the more SpaceX will acquire.

What are these chances?

Satellite imagery is outdated and lacks detail. We can improve greatly. Synthetic aperture radar can take beautiful shots like this:

Have you ever used Google Maps and thought, "I want to see this in more detail"? What if I could view Earth live? What if we could livestream an infrared image of Earth?

We could launch hundreds of satellites with such mind-blowing visual precision of the Earth that we would dramatically improve the accuracy of our meteorological models; our agriculture; where crime is happening; where poachers are operating in the savannah; climate change; and who is moving military personnel where. Is that useful?

What if we could see Earth in real time? That affects businesses? That changes society?

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.