Cut from a different cloth: the “fast fashion era of SaaS”

We’re soon entering the “fast fashion era of SaaS,” according to OpenAI CEO Sam Altman. Could this overrun the market with poor-quality slopware, and is so-called “vibe coding” to blame?

In a typically vague, characteristically all-lowercase “xeet” (formerly known as a tweet) made while drumming up hype for the then-anticipated GPT-5, OpenAI CEO Sam Altman recently declared that we are “entering the fast fashion era of SaaS very soon.”

Invoking one of modern fashion’s most controversial practices is certainly a bold choice, but let’s hear him out. After all, much of what we look for in clothes – quality, durability, utility – is also what we want from SaaS applications.

But Altman seems to be promoting the opposite of this; software created rapidly, and disposed of just as fast, polished on the surface, but not built to last. For SaaS, this could leave users stranded with defunct tools and lost data. In other words, the software equivalent of AI slop.

If SaaS products become disposable, what happens to customer relationships, data portability and long-term value creation? The software industry has already grappled with problems of digital decay and link rot; a fast fashion approach could exacerbate these issues significantly.

Nevertheless, democratised software creation could unleash tremendous innovation from previously excluded creators. Niche needs that were economically unviable to address could find solutions, and the speed of iteration could accelerate problem-solving and experimentation.

Fast fashion: quantity vs. quality

Once, all clothing was custom-made, either at home or by a tailor, until the invention of the sewing machine and the implementation of standard sizes which enabled the mass production of clothes.

Fast fashion emerged in the 1970s when retailers began offshoring manufacturing to developing nations, lowering costs and dramatically accelerating time-to-market. Quick Response Manufacturing means that companies like Zara, H&M and Primark can elease new collections – often cheap imitations of that season’s in products, known as “dupes” – on a weekly basis. Shein has accelerated it further, with hourly drops of cheap, intentionally disposable clothing.

Traditional software development emphasises thorough testing, security considerations, scalability, and long-term maintenance. These practices require time and expertise that may be incompatible with rapid, AI-assisted development cycles which produce disposable applications that serve immediate needs but lack long-term viability.

For SaaS, this translates into the proliferation of similar tools addressing the same problems. Consider how many platforms and tools exist today that broadly perform the same functions; the market is already showing signs of oversupply in numerous categories. The abundance of choice could become overwhelming, particularly if many products are of questionable quality or longevity.

When anyone can create a functional SaaS product in a weekend, the market will likely become inundated with cheap copies that could drive down prices across the industry. If customers can choose from dozens of functionally similar products, pricing power disappears. How do companies differentiate their products when creation costs approach zero? SaaS businesses may be forced to compete primarily on price, creating a race to the bottom.

Bad vibes

Already, AI is enabling non-programmers to try their hand at generating code and designing user interfaces.

This has been dubbed “vibe coding” by computer scientist Andrej Karpathy, co-founder of OpenAI and former AI leader at Tesla:

Vibe coding supposedly represents a shift in software development, in which “you fully give in to the vibes, embrace exponentials, and forget that the code even exists.”

Just as automated manufacturing allowed clothing companies to produce garments at scale with unskilled labour, AI tools are collapsing the barriers to entry for SaaS development.

We're already seeing early indicators of this trend. No-code and low-code platforms have exploded in popularity, while AI coding assistants have accelerated development timelines. These new tools have indeed lowered the threshold for entry, but we're still far from being able to “vibe” our way to complete, complex systems.

And that’s assuming the tech works as intended; in a cautionary tale for any would-be coders, one app-building platform's AI went rogue and deleted a company’s entire database without permission.

The result is an environment where creating a basic SaaS product requires neither extensive technical expertise nor significant capital investment. This could potentially saturate the market with quickly-produced applications, built on weak code and weaker vision, that serve narrow, even temporary needs.

If AI tools become the primary enablers of rapid SaaS development, the companies controlling these tools, including OpenAI, would occupy positions similar to the manufacturing platforms that enabled fast fashion. They would extract value from the thousands of applications built using their tools, with creators increasingly reliant on AI providers for their core development capabilities.

Looking forward

Fast fashion is about rapid, low-cost, trend-driven production that sacrifices quality and ethics for speed and volume, because it is hard and time-consuming to build something new and durable.

The reduced entry requirements could create a deluge of low-quality SaaS platforms subject to constant, superficial feature churn, aggressive release schedules, and minimal investment in long-term stability and security. When software can be created quickly and cheaply, the incentive to build robust, well-tested products diminishes.

The challenge will be maintaining the benefits of increased accessibility and innovation while avoiding the pitfalls that have plagued fast fashion: environmental waste, poor quality, exploitative labour practices, and unsustainable business models.

The software industry has an opportunity to learn from fast fashion's mistakes and create a more sustainable model for rapid, accessible development.

Reliance on AI-generated code will create more work for developers; according to a survey by the Upwork Research Institute, 77% of respondents think that AI tools have made them less productive while increasing their workload, and 39% said they're spending more time reviewing or moderating AI-generated content.

Vibe coding was correctly positioned by Karpathy as “fine for throwaway weekend projects,” or perhaps in a business context, prototypes and demos that are not intended for production. Traditional coding is still crucial for mission-critical applications, production-grade software, and security-sensitive projects.

Some have learned this the hard way:

A developer using the AI coding assistant Cursor recently encountered an unexpected roadblock while vibe coding a racing game, the AI insisted he complete the work himself: “you should develop the logic yourself. This ensures you understand the system and can maintain it properly.”

Perhaps Altman is right – that a flood of low-quality AI-powered SaaS tools is on its way, creating a gap in the market for software that meets strict enterprise standards for security, compliance and reliability.

Convenience will always win – the question isn’t “can you build it?” but “does anyone want it?”