Nach der Kündigung arbeitete ein Programmierer 300 Tage lang wie besessen an der Entwicklung eines KI - Tools und investierte 20.000 US - Dollar, doch es kamen 0 Nutzer und 0 Einnahmen. Eine Warnung aus dem finanziellen Desaster eines Programmierers

"He quit his high - paying job as an architect, worked day and night for nearly 10 months, invested his only $20,000 savings, and worked himself bald to create an AI design tool for his startup. He fantasized that users would flock to it after its launch. Little did he know that reality gave him a heavy blow: in the end, he earned $0 and had 0 users."

These words are heart - wrenching, yet they are the truth that this senior engineer personally encountered on his AI startup journey.

When he posted this experience on Reddit, it immediately sparked a heated discussion. The comment section was filled with developers and independent entrepreneurs who felt like "fellow sufferers." Many people sighed helplessly: They clearly followed the advice of OpenAI CEO Sam Altman and didn't create another ChatGPT, but still couldn't escape the fate of a solo startup. So, where exactly did the problem lie?

From a High - Paying Architect to a Desperate AI Entrepreneur

This user is named Sorry - Bat - 9609. He is from Silicon Valley and has worked as an engineering architect at Walmart, Visa, and Target, with over 15 years of software engineering experience.

With a passion for AI and design, he resolutely quit his job and embarked on entrepreneurship, aiming to develop a SaaS product.

The idea of starting a business came to him partly because of the arrival of the AI era and partly because he was already tired of the old and templated infographic tools in 2017. He even directly named them, saying, "I'm referring to tools like Canva and Visme. It takes 3 hours to create a mediocre graph. It's simply a waste of time."

So, he developed a tool called InfographsAI, an AI - driven platform that can quickly generate unique, client - level infographics, attempting to revolutionize those infographic products that users have long complained about as outdated.

"I spent 10 months writing the code. I think this is an innovation in the AI design industry. It doesn't rely on templates and doesn't require manual editing. It can generate 100% unique designs in real - time based on your theme every time."

Judging from the function description, InfographsAI is indeed quite impressive:

Instant generation: Input any content, whether it's a YouTube video link, PDF, website, image, text, or data file (some are still in testing). The AI can extract information, automatically typeset, match fonts and colors, and generate charts and visual content - the whole process takes no more than 200 seconds.

Completely abandon templates: Each design is customized according to your content, no longer the "one - size - fits - all" template content like Canva.

6 artistic styles: Cartoon, Pixar, Anime, Fairy Tale, Vector, Minimalist - making your visual content stand out.

Intelligent functions: Automatically verify facts to avoid data errors, generate charts in real - time with SmartGraph, and you can choose between "normal tone" or "sarcastic tone."

Global adaptation: Supports over 35 languages and has cultural context awareness.

Doesn't it sound quite good?

It's worth mentioning that its working principle is also very simple:

Upload any content

Select the style and tone

Generate a publishable infographic within seconds - no design skills are required at all

When he just finished the development, Sorry - Bat - 960 was full of confidence. After all, he spent nearly 10 months continuously improving the functions, constantly adding "one more cool function" as his daily routine. He wanted to integrate all his creativity into this product as perfectly as possible.

But when he finally pressed the "Publish" button, he found that: The world was completely silent. There were no users, no revenue, and even no one knew this tool existed.

He said helplessly:

"Even the few people who accidentally tried it thought it was good - but how did they discover it? It was like a coincidence."

"Perfection Lost to Lack of Users"

This disastrous startup experiment not only completely shattered his fantasy but also uncovered a frequently overlooked scar in the AI startup circle: those projects that seem "technologically awesome" often fail in the face of the fact that no one needs them.

Especially when engineers rush into the battlefield with boundless enthusiasm for technology but ignore users, needs, and promotion, they often repeat the same failure script.

During the post - mortem, Sorry - Bat - 9609 summarized the core mistakes he made in this painful experience - "These mistakes almost crushed me":

No demand validation: He only created what he thought was cool and never asked anyone "Would you use this?"

Function stacking: He spent 8 months continuously adding "one last cool function," and finally it became a bloated prototype.

The perfectionism trap: He spent several weeks optimizing the code and design that no one saw, and he became his own biggest enemy.

Zero marketing: He naively believed that "if it's good, users will naturally come." In the end, they didn't.

Out of touch with reality: He didn't research whether others were already solving similar problems. If you want to create a "disruptive" product or start a new category, you have to let people know it exists first. And no one would expect that "infographics can be done without templates."

As a result, as mentioned above, after spending more than 300 days on InfographsAI, he earned $0, lost countless nights and weekends, and his $20,000 savings.

The Programmer's Pursuit of Technical Perfection Doesn't Apply to Products

Although his startup failed miserably, Sorry - Bat - 9609 said that he wouldn't give up. He was glad that he at least created the product - 'Some testers gave very positive feedback. One said it converted a "dense PDF" into a Pixar - style illustration in 180 seconds. This is also a starting point.'

Meanwhile, he is also trying to change his way of thinking. He said: "The biggest mistake was that I always approached things from an engineer's perspective, thinking about making things right, beautiful, and of high quality. But entrepreneurship requires a different kind of thinking - you have to let people know what you're doing first."

Stop thinking like a programmer ("How can I make it more perfect?") and start thinking like a marketer ("How can I make others care about this product?").

Start "building a community and finding an audience" before you actually start writing code. Start talking about your product as early as possible.

No one told me how easy it is to spend 10 months developing something in silence while no one knows what you're doing.

Based on these thoughts, he started taking action:

Verify demand early: Talk to 10 - 20 potential users before writing code.

Launch the MVP quickly: Don't spend 10 months. Finish the core functions in 6 weeks and launch it.

Do marketing from day one: Build an audience while writing code.

Focus on user feedback: Let people use the product before adding more functions.

Set a launch deadline: Deliver it even if it's not perfect. Don't delay for too long.

What's the Possibility of Replacing Canva?

Through this, Sorry - Bat - 9609 hopes that his experience can ring an alarm bell for more AI entrepreneurs and prevent others from falling into the same startup pitfalls.

When talking about future plans and someone asked about the use of this tool, Sorry - Bat - 9609 replied:

"I hope it can replace Canva - that 'ancestor' design tool that requires manual operation for everything. I'm creating a new market: just click once, no design knowledge or editing skills are required.

Today, it takes 1 hour to generate visual content from data manually. My tool has the potential to replace that process."

In the comment section, another netizen named Business_Young416 also put forward a calm and sharp suggestion. He said:

"Although I can understand what you're doing, as a person who specializes in PMF (Product - Market Fit) research, I have some ideas to share:

Canva is almost all - purpose. It's a one - stop tool that users are already very familiar with and can almost meet all usage scenarios. The challenge you're facing is like all graphic editing tools trying to shake Adobe or Figma, or like a new language trying to convince developers that 'I'm better than JavaScript.'

It will be very difficult to convince content creators to pay for a professional tool that only solves a single problem, given the numerous existing options. There are also tools like Gamma.app that are doing quite well. I'm a content creator myself, and I even find it hard to convince myself to pay for this tool.

How many infographics does a business really need to justify paying for this tool continuously?

You may need to consider adjusting the pricing model or finding a vertical market that has a real and large - scale demand for infographics (the medical industry? Not sure), and where Canva doesn't perform well enough in those scenarios.

Advertising is not for directly selling products but for finding your market positioning. Try different titles, different focuses, and different industry directions. Once you find the right entry point, focus on it and use it to guide subsequent feature development."

In addition, many peers encouraged him:

"Good job. Keep going. You're just at the beginning. For me, it took 2 years, 4 stints of 12 - 16 - hour workdays, 7 days a week for my startup. Find a good co - founder, talk to users, and take action. Good luck."

"The road to success is paved with failures. I hope you can apply what you've learned to your next attempt. Most people don't achieve anything. So, no matter how you perceive success, you're far ahead of most people."

This also makes more practitioners deeply realize that entrepreneurship has never been an easy road: an AI product can't succeed just with code and cool functions. Finding the real market and understanding the users' language are the keys for a product to survive.

What do you think of Sorry - Bat - 9609's experience?

Reference: https://www.reddit.com/r/SaaS/comments/1lg2zw7/i_left_my_job_spent_almost_10_months_20k_of_loss/

This article is from the WeChat official account "CSDN", compiled by Tu Min, and published by 36Kr with authorization.