Dharmesh Shah co-founded HubSpot. WordPlay reached $90,000 per month in revenue without utilizing any of his wealth.

His method:

Take Advantage Of An Established Trend

Remember Wordle? Dharmesh was instantly hooked. As was the tech world.

HubSpot's co-founder noted inefficiencies in a recent My First Million episode. He wanted to play daily. Dharmesh, a tinkerer and software engineer, decided to design a word game.

He's a billionaire. How could he?

1. Wordle had limitations in his opinion;

2. Dharmesh is fundamentally a developer. He desired to start something new and increase his programming knowledge;

3. This project may serve as an excellent illustration for his son, who had begun learning about software development.

Better It Up

Building a new Wordle wasn't successful.

WordPlay lets you play with friends and family. You could challenge them and compare the results. It is a built-in growth tool.

WordPlay features:

• the capacity to follow sophisticated statistics after creating an account;

• continuous feedback on your performance;

• Outstanding domain name (wordplay.com).

Project Development

WordPlay has 9.5 million visitors and 45 million games played since February.

HubSpot co-founder credits tremendous growth to flywheel marketing, pushing the game through his own following.

Choosing an exploding specialty and making sharing easy also helped.

Shah enabled Google Ads on the website to test earning potential. Monthly revenue was $90,000.

That's just Google Ads. If monetization was the goal, a specialized ad network like Ezoic could double or triple the amount.

Wordle was a great buy for The New York Times at $1 million.

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.

People don’t believe in influencers anymore; they need people like you.

Social media influencers have dominated for years. We've seen videos, images, and articles of *famous* individuals unwrapping, reviewing, and endorsing things.

This industry generates billions. This year, marketers spent $2.23 billion on Instagram, $1 million on Youtube, and $775 million on Tiktok. This marketing has helped start certain companies.

Influencers are dying, so ordinary people like us may take over this billion-dollar sector. Why?

Why influencers are perishing

Most influencers lie to their fans, especially on Instagram. Influencers' first purpose was to make their lives so flawless that others would want to buy their stuff.

In 2015, an Australian influencer with 600,000 followers went viral for revealing all her photos and everything she did to seem great before deleting her account.

“I dramatically edited the pictures, I manipulated the environements, and made my life look perfect in social media… I remember I obsessively checked the like count for a full week since uploading it, a selfie that now has close to 2,500 likes. It got 5 likes. This was when I was so hungry for social media validation … This was the reason why I quit social media: for me, personally, it consumed me. I wasn’t living in a 3D world.”

Influencers then lost credibility.

Influencers seem to live in a bubble, separate from us. Thanks to self-popularity love's and constant awareness campaigns, people find these people ridiculous.

Influencers are praised more for showing themselves as natural and common than for showing luxuries and lies.

Little by little, they are dying, making room for a new group to take advantage of this multi-million dollar business, which gives us (ordinary people) a big opportunity to grow on any content creation platform we want.

Why this is your chance to develop on any platform for creating content

In 2021, I wrote “Not everyone who talks about money is a Financial Advisor, be careful of who you take advice from,”. In it, I warned that not everyone with a large following is a reputable source of financial advice.

Other writers hated this post and said I was wrong.

People don't want Jeff Bezos or Elon Musk's counsel, they said. They prefer to hear about their neighbor's restroom problems or his closest friend's terrible business.

Real advice from regular folks.

And I found this was true when I returned to my independent YouTube channel and had more than 1000 followers after having abandoned it with fewer than 30 videos in 2021 since there were already many personal finance and travel channels and I thought mine wasn't special.

People appreciated my videos because I was a 20-something girl trying to make money online, and they believed my advice more than that of influencers with thousands of followers.

I think today is the greatest time to grow on any platform as an ordinary person. Normal individuals give honest recommendations about what works for them and look easier to make because they have the same options as us.

Nobody cares how a millionaire acquired a Lamborghini unless it's entertaining. Education works now. Real counsel from average people is replicable.

Many individuals don't appreciate how false influencers seem (unreal bodies and excessive surgery and retouching) since it makes them feel uneasy.

That's why body-positive advertisements have been so effective, but they've lost ground in places like Tiktok, where the audience wants more content from everyday people than influencers living amazing lives. More people will relate to your content if you appear genuine.

Last thoughts

Influencers are dwindling. People want more real people to give real advice and demonstrate an ordinary life.

People will enjoy anything you tell about your daily life as long as you provide value, and you can build a following rapidly if you're honest.

This is a millionaire industry that is getting more expensive and will go with what works, so stand out immediately.

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.