Startup entrepreneurs assume what risks when pitching?

Last week I signed four NDAs.

Four!

NDA stands for non-disclosure agreement. A legal document given to someone receiving confidential information. By signing, the person pledges not to share the information for a certain time. If they do, they may be in breach of contract and face legal action.

Companies use NDAs to protect trade secrets and confidential internal information from employees and contractors. Appropriate. If you manage a huge, successful firm, you don't want your employees selling their information to your competitors. To be true, business NDAs don't always prevent corporate espionage, but they usually make employees and contractors think twice before sharing.

I understand employee and contractor NDAs, but I wasn't asked to sign one. I counsel entrepreneurs, thus the NDAs I signed last week were from startups that wanted my feedback on their concepts.

I’m not a startup investor. I give startup guidance online. Despite that, four entrepreneurs thought their company ideas were so important they wanted me to sign a generically written legal form they probably acquired from a shady, spam-filled legal templates website before we could chat.

False. One company tried to get me to sign their NDA a few days after our conversation. I gently rejected, but their tenacity encouraged me. I considered sending retroactive NDAs to everyone I've ever talked to about one of my startups in case they establish a successful company based on something I said.

Two of the other three NDAs were from nearly identical companies. Good thing I didn't sign an NDA for the first one, else they may have sued me for talking to the second one as though I control the firms people pitch me.

I wasn't talking to the fourth NDA company. Instead, I received an unsolicited email from someone who wanted comments on their fundraising pitch deck but required me to sign an NDA before sending it.

That's right, before I could read a random Internet stranger's unsolicited pitch deck, I had to sign his NDA, potentially limiting my ability to discuss what was in it.

You should understand. Advisors, mentors, investors, etc. talk to hundreds of businesses each year. They cannot manage all the companies they deal with, thus they cannot risk legal trouble by talking to someone. Well, if I signed NDAs for all the startups I spoke with, half of the 300+ articles I've written on Medium over the past several years could get me sued into the next century because I've undoubtedly addressed topics in my articles that I discussed with them.

The four NDAs I received last week are part of a recent trend of entrepreneurs sending out NDAs before meetings, despite the practical and legal issues. They act like asking someone to sign away their right to talk about all they see and hear in a day is as straightforward as asking for a glass of water.

Given this inflow of NDAs, I wanted to briefly remind entrepreneurs reading this blog about the merits and cons of requesting investors (or others in the startup ecosystem) to sign your NDA.

Benefits of having investors sign your NDA include:

None. Zero. Nothing.

Disadvantages of requesting investor NDAs:

• You'll come off as an amateur who has no idea what it takes to launch a successful firm.

• Investors won't trust you with their money since you appear to be a complete amateur.

• Printing NDAs will be a waste of paper because no genuine entrepreneur will ever sign one.

I apologize for missing any cons. Please leave your remarks.

SaaS spending is 40/40/20. 40% of operating expenses should be R&D, 40% sales and marketing, and 20% G&A. We wanted to see the statistics behind the rules of thumb. Since October 2017, 73 SaaS startups have gone public. Perhaps the rule of thumb should be 30/50/20. The data is below.

30/50/20. R&D accounts for 26% of opex, sales and marketing 48%, and G&A 22%. We think R&D/S&M/G&A should be 30/50/20.

There are outliers. There are exceptions to rules of thumb. Dropbox spent 45% on R&D whereas Zoom spent 13%. Zoom spent 73% on S&M, Dropbox 37%, and Bill.com 28%. Snowflake spent 130% of revenue on S&M, while their EBITDA margin is -192%.

G&A shouldn't stand out. Minimize G&A spending. Priorities should be product development and sales. Cloudflare, Sendgrid, Snowflake, and Palantir spend 36%, 34%, 37%, and 43% on G&A.

Another myth is that COGS is 20% of revenue. Median and averages are 29%.

Where is the profitability? Data-driven operating income calculations were simplified (Revenue COGS R&D S&M G&A). 20 of 73 IPO businesses reported operational income. Median and average operating income margins are -21% and -27%.

As long as you're growing fast, have outstanding retention, and marquee clients, you can burn cash since recurring income that doesn't churn is a valuable annuity.

The data was compelling overall. 30/50/20 is the new 40/40/20 for more established SaaS enterprises, unprofitability is alright as long as your business is expanding, and COGS can be somewhat more than 20% of revenue.

Failures inspire young developers

Jason citron created many products.

These products flopped.

Microsoft offered $10 billion for one of these products.

He rejected the offer since he was so confident in his success.

Let’s find out how he built a product that is currently valued at $15 billion.

Early in his youth, Jason began learning to code.

Jason's father taught him programming and IT.

His father wanted to help him earn money when he needed it.

Jason created video games and websites in high school.

Jason realized early on that his IT and programming skills could make him money.

Jason's parents misjudged his aptitude for programming.

Jason frequented online programming communities.

He looked for web developers. He created websites for those people.

His parents suspected Jason sold drugs online. When he said he used programming to make money, they were shocked.

They helped him set up a PayPal account.

Florida higher education to study video game creation

Jason never attended an expensive university.

He studied game design in Florida.

“Higher Education is an interesting part of society… When I work with people, the school they went to never comes up… only thing that matters is what can you do…At the end of the day, the beauty of silicon valley is that if you have a great idea and you can bring it to the life, you can convince a total stranger to give you money and join your project… This notion that you have to go to a great school didn’t end up being a thing for me.”

Jason's life was altered by Steve Jobs' keynote address.

After graduating, Jason joined an incubator.

Jason created a video-dating site first.

Bad idea.

Nobody wanted to use it when it was released, so they shut it down.

He made a multiplayer game.

It was released on Bebo. 10,000 people played it.

When Steve Jobs unveiled the Apple app store, he stopped playing.

The introduction of the app store resembled that of a new gaming console.

Jason's life altered after Steve Jobs' 2008 address.

“Whenever a new video game console is launched, that’s the opportunity for a new video game studio to get started, it’s because there aren’t too many games available…When a new PlayStation comes out, since it’s a new system, there’s only a handful of titles available… If you can be a launch title you can get a lot of distribution.”

Apple's app store provided a chance to start a video game company.

They released an app after 5 months of work.

Aurora Feint is the game.

Jason believed 1000 players in a week would be wonderful. A thousand players joined in the first hour.

Over time, Aurora Feints' game didn't gain traction. They don't make enough money to keep playing.

They could only make enough for one month.

Instead of buying video games, buy technology

Jason saw that they established a leaderboard, chat rooms, and multiplayer capabilities and believed other developers would want to use these.

They opted to sell the prior game's technology.

OpenFeint.

Assisting other game developers

They had no money in the bank to create everything needed to make the technology user-friendly.

Jason and Daniel designed a website saying:

“If you’re making a video game and want to have a drop in multiplayer support, you can use our system”

TechCrunch covered their website launch, and they gained a few hundred mailing list subscribers.

They raised seed funding with the mailing list.

Nearly all iPhone game developers started adopting the Open Feint logo.

“It was pretty wild… It was really like a whole social platform for people to play with their friends.”

What kind of a business model was it?

OpenFeint originally planned to make the software free for all games. As the game gained popularity, they demanded payment.

They later concluded it wasn't a good business concept.

It became free eventually.

Acquired for $104 million

Open Feint's users and employees grew tremendously.

GREE bought OpenFeint for $104 million in April 2011.

GREE initially committed to helping Jason and his team build a fantastic company.

Three or four months after the acquisition, Jason recognized they had a different vision.

He quit.

Jason's Original Vision for the iPad

Jason focused on distribution in 2012 to help businesses stand out.

The iPad market and user base were growing tremendously.

Jason said the iPad may replace mobile gadgets.

iPad gamers behaved differently than mobile gamers.

People sat longer and experienced more using an iPad.

“The idea I had was what if we built a gaming business that was more like traditional video games but played on tablets as opposed to some kind of mobile game that I’ve been doing before.”

Unexpected insight after researching the video game industry

Jason learned from studying the gaming industry that long-standing companies had advantages beyond a single release.

Previously, long-standing video game firms had their own distribution system. This distribution strategy could buffer time between successful titles.

Sony, Microsoft, and Valve all have gaming consoles and online stores.

So he built a distribution system.

He created a group chat app for gamers.

He envisioned a team-based multiplayer game with text and voice interaction.

His objective was to develop a communication network, release more games, and start a game distribution business.

Remaking the video game League of Legends

Jason and his crew reimagined a League of Legends game mode for 12-inch glass.

They adapted the game for tablets.

League of Legends was PC-only.

So they rebuilt it.

They overhauled the game and included native mobile experiences to stand out.

Hammer and Chisel was the company's name.

18 people worked on the game.

The game was funded. The game took 2.5 years to make.

Was the game a success?

July 2014 marked the game's release. The team's hopes were dashed.

Critics initially praised the game.

Initial installation was widespread.

The game failed.

As time passed, the team realized iPad gaming wouldn't increase much and mobile would win.

Jason was given a fresh idea by Stan Vishnevskiy.

Stan Vishnevskiy was a corporate engineer.

He told Jason about his plan to design a communication app without a game.

This concept seeded modern strife.

“The insight that he really had was to put a couple of dots together… we’re seeing our customers communicating around our own game with all these different apps and also ourselves when we’re playing on PC… We should solve that problem directly rather than needing to build a new game…we should start making it on PC.”

So began Discord.

Online socializing with pals was the newest trend.

Jason grew up playing video games with his friends.

He never played outside.

Jason had many great moments playing video games with his closest buddy, wife, and brother.

Discord was about providing a location for you and your group to speak and hang out.

Like a private cafe, bedroom, or living room.

Discord was developed for you and your friends on computers and phones.

You can quickly call your buddies during a game to conduct a conference call. Put the call on speaker and talk while playing.

Discord wanted to give every player a unique experience. Because coordinating across apps was a headache.

The entire team started concentrating on Discord.

Jason decided Hammer and Chisel would focus on their chat app.

Jason didn't want to make a video game.

How Discord attracted the appropriate attention

During the first five months, the entire team worked on the game and got feedback from friends.

This ensures product improvement. As a result, some teammates' buddies started utilizing Discord.

The team knew it would become something, but the result was buggy. App occasionally crashed.

Jason persuaded a gamer friend to write on Reddit about the software.

New people would find Discord. Why not?

Reddit users discovered Discord and 50 started using it frequently.

Discord was launched.

Rejecting the $10 billion acquisition proposal

Discord has increased in recent years.

It sends billions of messages.

Discord's users aren't tracked. They're privacy-focused.

Purchase offer

Covid boosted Discord's user base.

Weekly, billions of messages were transmitted.

Microsoft offered $10 billion for Discord in 2021.

Jason sold Open Feint for $104m in 2011.

This time, he believed in the product so much that he rejected Microsoft's offer.

“I was talking to some people in the team about which way we could go… The good thing was that most of the team wanted to continue building.”

Last time, Discord was valued at $15 billion.

Discord raised money on March 12, 2022.

The $15 billion corporation raised $500 million in 2021.

This is how to build a car while driving.

Building a high-growth startup is compared to building a car while it's speeding down the highway.

How to plan without going crazy? Or, without losing team, board, and investor buy-in?

I just delivered our company's product roadmap for the rest of the year. Complete. Thorough. Page-long. I'm optimistic about its chances of surviving as everything around us changes, from internal priorities to the global economy.

It's tricky. This isn't the first time I've created a startup roadmap. I didn't invent a document. It took time to deliver a document that will be relevant for months.

Goals matter.

Although they never change, goals are rarely understood.

This is the third in a series about a startup's unique roadmapping needs. Velocity is the intensity at which a startup must produce to survive.

A high-growth startup moves at breakneck speed, which I alluded to when I said priorities and economic factors can change daily or weekly.

At that speed, a startup's roadmap must be flexible, bend but not break, and be brief and to the point. I can't tell you how many startups and large companies develop a product roadmap every quarter and then tuck it away.

Big, wealthy companies can do this. It's suicide for a startup.

The drawer thing happens because startup product roadmaps are often valid for a short time. The roadmap is a random list of features prioritized by different company factions and unrelated to company goals.

It's not because the goals changed that a roadmap is shelved or ignored. Because the company's goals were never communicated or documented in the context of its product.

In the previous post, I discussed how to turn company goals into a product roadmap. In this post, I'll show you how to make a one-page startup roadmap.

In a future post, I'll show you how to follow this roadmap. This roadmap helps you track company goals, something a roadmap must do.

Be vague for growth, but direct for execution.

Here's my plan. The real one has more entries and more content in each.

Let's discuss smaller boxes.

Product developers and engineers know that the further out they predict, the more wrong they'll be. When developing the product roadmap, this rule is ignored. Then it bites us three, six, or nine months later when we haven't even started.

Why do we put everything in a product roadmap like a project plan?

Yes, I know. We use it when the product roadmap isn't goal-based.

A goal-based roadmap begins with a document that outlines each goal's idea, execution, growth, and refinement.

Once the goals are broken down into epics, initiatives, projects, and programs, only the idea and execution phases should be modeled. Any goal growth or refinement items should be vague and loosely mapped.

Why? First, any idea or execution-phase goal will result in growth initiatives that are unimaginable today. Second, internal priorities and external factors will change, but the goals won't. Locking items into calendar slots reduces flexibility and forces deviation from the single source of truth.

No soothsayers. Predicting the future is pointless; just prepare.

A map is useless if you don't know where you're going.

As we speed down the road, the car and the road will change. Goals define the destination.

This quarter and next quarter's roadmap should be set. After that, you should track destination milestones, not how to get there.

When you do that, even the most critical investors will understand the roadmap and buy in. When you track progress at the end of the quarter and revise your roadmap, the destination won't change.

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.