One of the most important questions of the AI era is also one of the most personal: what happens to our jobs? Most takes online are surface-level truisms like “tasks will be automated” or “AI agents will slowly replace human labor.” The few that try to go deeper tend to fall under one of two headlines:

1: “We’ll see $1B companies run by one person.”

Euphoric, founder-core speculation. Not necessarily wrong– the more broadly applicable kernel of truth here is that solopreneurs and small teams will be more capable than ever before. But this particular expression of that truth is useless for 99.9999% of people.

2: “AI will replace your job, your husband’s job, your baby daddy’s job, and all your future offsprings’ jobs forever and ever.”

Paints all working humans as helpless victims. Utterly, deeply unhelpful.

I want to propose a third headline. One that I see unfolding in real time. One that’s helpful. And, coincidentally, one that’s highly optimistic:

New, third, headline:

AI is unlocking a new paradigm of employment: One where we don’t have “jobs”, but we direct AI at what we’re passionate about to generate small, independent, sustainable self-supporting cash flow.

Let me explain.

In late March 2025, my friend Eric had the idea for an iPhone-shaped slab of acrylic that could be used as a sort of “placebo” to help people reduce their dependency on their phones. The same day he had the idea, he went to ChatGPT and asked it to create an image of what this clear, phone-shaped device would look like. After seeing his idea visually come to life, he used ChatGPT to brainstorm copy for the kickstarter campaign, create a logo, research materials, find manufacturers, and draft an order request. Within 10 days of his first message to ChatGPT, he had a prototype in hand. (He mailed one to me and I posted a video of me scrolling on it while standing in line at a boba shop. Maybe you’ve seen it– it’s racked up over 150 million views so far.) 

Around the same time in early March, my friend, Sirio, was figuring out a way to make AI generated images that were truly photoreal. (ChatGPT Image generation looks amazing. Sirio’s are even better.) Once he got it working, he quickly realized the business opportunity– he just needed to figure out how to turn his duct-taped home-brewed solution into an actual product. Sirio, who is not a software engineer and has never written a line of code in his life, built a prototype of his product using Bolt.new and Cursor, two popular “vibe coding” platforms that take product descriptions in normal English and turn them into real, working code. He hired two engineers in Sri Lanka to add the final touches, and launched it on his Instagram and YouTube channels on April 5– just three weeks after his first vibecoding session. Less than 90 days after the launch, the app is bringing in $89K per month. 

In both of these examples, smart, curious people, with a strong bias towards action and a knowledge of basic AI tools went from an “idea in their head” to using AI as their co-conspirator to actually make it real in a matter of weeks. And, in both of these cases, they walked away with money in their pocket. 

Eric and Sirio are the tip of the iceberg on a seismic shift towards sustainable, independent, cash-flow projects built with AI, and distributed through social media. The conditions are aligning perfectly for a storm of regular, everyday people to start using AI to make stuff that makes money. First, the cost to start something is lower than ever: you don’t need $150K and a dev team– you need $100 and a weekend. Second, the technical barriers are dissolving. AI has turned into an endlessly patient cofounder with answers to almost anything. Third, social media provides built-in distribution: creators don’t need agencies or ad budgets– they need a story and a phone. Fourth, Gen Z is already leaning in. Surveys show a majority of them not only want to build side hustles– they’re already doing it. And finally, to be blunt, it’s possible many people won’t have a choice. 

Let’s dive into each of these conditions deeper.

Condition #1: AI lowers the risk to starting something

The best word we have to describe Eric and Sirio is “entrepreneur”. But the definition doesn’t quite fit– see if you can tell why: An entrepreneur is “a person who organizes and operates a business or businesses, taking on greater than normal financial risks in order to do so.”

If you said “greater than normal financial risks”, congratulations. You win a cookie. 

Before AI, if you weren’t a software engineer and had a brilliant idea for an app, you would need $150K and a dev team to test it out. This meant that the conditions to spawn a new entrepreneur would be that they had: (1) startup capital or connections to secure startup capital, (2) a great ability to pitch themselves and their idea to investors, (3) radical self-belief, and (4) a wife that wouldn’t leave them if they quit their stable day job. 

Now, you can test an idea for $100 by yourself. The risk has lowered, and brought the risk threshold required for success down with it. This means an entirely new subset of the population who didn’t fit the risk profile of an entrepreneur before, could now take a swing at building something of their own.

To extend this even further, because AI allows you to spin up ideas quickly, it means you can throw out bad ideas and start over with new ones just as fast. If Uber for dogs doesn’t work, you’ve lost a month, not a year. Low ego people who don’t take the failures personally will compound learnings across all these experiments quickly, which will lead to them finding product-market-fit faster... And THAT will mean that these people stay entrepreneurs because their ideas will work. Nothing is more addictive to a builder than realizing you’ve created something that people want.

Condition #2: AI lowers the technical knowledge barrier to entrepreneurship. 

All of the parts of starting a business that feel scary and foreign the first time around: finding customers, building business plans, writing code… AI has expert-level knowledge on all of it. Anyone can talk to their LLM chatbot like the endlessly patient subject matter expert that it is– breaking down codebases, navigating legal contracts, validating business ideas, and everything in between. (Pro tip: Throw a nice little “Respond like Paul Graham” on there, and you’re cooking with gas.) 

For those who are specifically interested in bringing ideas to life with code– every single day in my comments and DMs I read stories of people with zero coding experience who are bringing their app ideas to life with vibecoding platforms like Bolt.new and Replit. The excuse of “I’m not a technical person, therefore I can’t start a startup” is dissolving quickly as word gets out that these tools can get you 80% of the way there for $20 in a week.

For people whose biggest blocker to being an “entrepreneur” in the past was that they felt unsure, awkward, or insecure about their technical abilities and lack of experience, AI will make them feel more confident because they know if they ever get stuck or feel unsure, they can simply ask AI for help– or literally just tell them exactly what to do. (Well-connected, high-status insiders will shrug this off as “hardly a contributing factor.” Moms in Illinois know what’s up.)

Condition #3: It’s not just AI happening right now. It’s AI PLUS social media.

An extremely, extremely, extremely important factor in the question of “Why now?” is social media. In the same way that AI democratizes people’s ability to turn their ideas into prototypes, social media democratizes people’s ability to share the stories behind their products with the world. It’s possible to imagine a world in which AI was invented at a time before Instagram. Think about it: If we had AI in the 90s, when information about new products was disseminated through magazine pages, cable TV ad spots, and… idk… billboards??-- Then these home-brewed products/services might die at the distribution phase, lest its creator secure funding, press, or an ad agency to find their first users. At the very least, they’d struggle to reach the far corners of the world as quickly and as effectively as our curated, algorithmic feeds can deliver. For all of its shortcomings, social media has lowered the barrier to entry for eyeballs to a level playing field for everyone– it costs $0 to get started, and we all start with 0 followers and 0 engagement. 

Also, the creator economy is growing like crazy. Goldman Sachs predicts the influencer/creator industry will double from ~$250 billion in 2023 to nearly $500 billion by 2027. This growth comes from both consumer demand for online content and heavy investment by tech platforms in monetization tools for creators. In 2026 an estimated 50% of marketing spend will go to creators. In practical terms, this means even small-scale creators can tap into income– and those incomes are trending upward. 

Condition #4: Gen Z wants to be entrepreneurs

The youngest cohort in the workforce isn’t afraid of working for themselves. They’re excited by it. Multiple surveys in 2024 and 2025 show that well over half of Gen Z harbors entrepreneurial ambitions. A global Samsung/Edelman survey found 73% of Gen Z professionals dream of starting a side hustle (and many already took steps in the past year). Similarly, Intuit’s 2024 report noted nearly two-thirds of 18–35 year-olds have started or plan to start a side venture, and 65% intend to carry these hustles forward into.

Combine this embrace of entrepreneurship with the other typical advantages of youth: kids have more time, less responsibilities, and can actually spend time learning where adults and companies are slow and resistant to change. Additionally, they don’t have preconceived notions of how things are done, and so they wield AI like an operating system and less like an add-on

With Gen Z embracing entrepreneurship, side hustles, and content creation at unprecedented rates, it’s easy to imagine their appetite for self-sufficiency pairing easily with AI’s ability to deliver on that promise.

Condition #5: Tbh, many people will have no other option than to become self-sufficient

This essay started soberly, and it’ll end that way too: Jobs are getting automated. AI is replacing humans. Companies are laying folks off. We’re halfway through 2025 and we’ve already seen nearly 700,000 job cuts– nearly the grand total for 2024. So the fifth and final reason for why we might see more people becoming self-reliant in the AI future is that they’ll simply have no other choice. If you want to put food on the table, you need cash flow. And if you can’t find the infrastructure to make it happen for you, you’ll have to create it on your own.

—

I believe we’re standing at the edge of a new era of self-sufficiency that doesn’t fit the current definition of entrepreneurship because it doesn’t demand the same all-consuming sacrifice. Soon, we’ll look back at the idea that you had to quit your job and raise a round just to test an idea and laugh. (Or sigh. Or both.) Of course, not everyone wants to work for themselves. But for those who do, or have fantasized about it but couldn’t afford to take the risk, the price of that risk just dropped. By a lot. The possibility for how a single person spends their hours, which comprise their days, which comprise their lives, is opening up. And it’s an ocean.

Catherine “CatGPT”

IG: @askcatgpt 

TikTok: @askcatgpt