Generative AI is a threat to SaaS companies. Here’s why.

• Generative AI comes with coding capabilities, making software development easier.
• This could disrupt how software is created, distributed and used, VCs and startup founders say.
• But the disappearance of traditional SaaS companies seems far away.

While ChatGPT is taking the world by storm, behind the scenes investors and technologists are talking about deeper disruptions to the inner workings of the existing software industry.

A new potential framework for software, dubbed “adaptable software” by researcher Philip Chernavsky, is where generative AI and humans work together to customize tools and even create entire applications. I am drawing a future to

The result will turn the traditional software industry upside down and call into question the value of SaaS companies in a world where ordinary people can build their own software.

“No code was the first step,” said Matt Turk, a partner at venture capital firm FirstMark. “This is the final chapter in the software eating away at the world, allowing more people to create enterprise software within their own companies.”

This would be a complete reversal for the industry. A software-as-a-service company, he’s been a disruptor, not a disruptor, for a decade. Investors are confident that subscription revenues will continue to grow steadily for years to come, which is why their valuations are so high. However, if generative AI really becomes widespread, that future may change dramatically.

Democratizing technology development

Venture capitalists and startup founders have been obsessed with the idea of ​​democratizing tech creation for years. This is a tag last year, as seen in the rise of low- and no-code startups like Airtable, which was last valued at $11 billion, and Webflow, which sold for $4 billion.

Building most software still required some technical knowledge. But now, the advent of generative AI tools like GitHub Copilot has opened up the ability to generate code using only natural language, said Ethan Kurzweil, a partner at Bessemer Venture Partners. told his Insider.

For Jake Saper, general partner at Emergence Capital, the first use cases that can be disrupted are simple, low-risk tasks and applications in small businesses. He said these cases are the least likely to disrupt business and require the least amount of coordination between companies.

Fika Ventures senior associate James Schecter says vertical software companies that take existing technology and make it easier to use in outdated industries also find their added value focused on convenience rather than actual product differentiation. It said it could be exposed to alternative threats.

Technicians are already using generative AI tools like Copilot to build simple apps like trivia games and Amazon discount sites.

Some late-stage tech startups are trying to stay ahead of the curve by sharing the power of creativity with their customers. One example is his AI writing assistant for knowledge-based startup Guru. This allows customers to create their own custom tones using generative AI. This challenges the conventional notion of software as a rigid tool with a fixed set of actions available to users, said Guru co-founder and CEO Rick Nucci. told Insider.

“We’ve been talking about ‘platforms’ in the SaaS world for a long time. This is the idea that someone creates a set of basic building blocks that can compose and shape what the customer wants. ” he said. “This is really a gradual change that’s happening.”

A new age of software

Some VCs and founders believe that generative AI has the potential to transform not only how we create technology, but also how we interact with it through hyper-personalization.

For example, new generative AI technology could help startups create customized user interfaces to each person’s exact preferences, said Bessemer partner Talia Goldberg. . She explained that ChatGPT is already showing signs of this by choosing to provide certain responses in the form of a table of data, even if the user didn’t specifically request it.

In more extreme cases, the entire tool could be AI-generated on the fly to replace common actions users take, said CRV head Brittany Walker.

Longer term, venture capitalists like NEA partner Vanessa Larco and investor Erad Gill believe that autonomous AI agents, rather than humans, will be the primary players interacting with software. One of her potential scenarios might be a world where an individual has a primary AI agent who coordinates and manages a number of “microagents” that can do everything from texting her to making dinner reservations. Maybe, Larco told her Insider.

This kind of connectivity and interaction—now the technical plumbing that makes various software programs work together—is critical business for many SaaS companies. What would happen to these SaaS businesses when generative AI models could automatically perform this work?

“Healthy Pressure” on Traditional SaaS Providers

Indeed, the disappearance of traditional software companies seems far away.

First, the choice between building software yourself or buying it from a third party has a significant opportunity cost.

“You don’t necessarily have to spend 10, 12, 15 hours sitting in front of a computer developing something if you can find something out of the box,” said CRV’s Walker. “The barriers would have to be very low for the necessary number of people to start writing their own bespoke software.”

Additionally, by paying an outside software vendor, people are putting the burden of safety, maintenance and accountability on to a third party, Emergence Capital’s Saper said.

But even skeptics admit that the threat of generative AI to traditional SaaS will require incumbent software companies to prove their worth.

“It’s theoretically easier, so the ‘build’ decision can be more attractive, so there’s probably a healthy amount of pressure,” Saper said. “There will be pressure on software vendors to deliver real value.”

Any hints? Contact Stephanie Palazzolo using your non-work device on the encrypted messaging app Signal (+1 979-599-8091), email (spalazzolo@insider.com), or Twitter DM. @steph_palazzolo.