(Bloomberg) – Amazon.com Inc. spent years hiring dominance and then hiring more people to manage all of them. CEO Andy Jassy is now trying to tame the bureaucracy of the creation, believing that the company needs to regain its agility and be more like the previous startups.
But what if Amazon is established today? As Jassy describes, its top 10 employees are not just “really smart, motivated, creative, ambitious,” but AI Power users are equipped with tools that can handle customer service, scheduling, and product prototypes alone? Can companies scale up faster and have fewer employees?
It’s the basic premise for a new series of startups to call themselves “AI natives” – companies built from AI not only in products, but also at the heart of workflows and team structures.
Jared Spataro of Microsoft Corp. is the chief marketing officer for AI work and for this reason has been researching Tiny, AI-Native. “They took a step back and said, what is the investment in our company? What is the output? What can we do with AI in the center?” he said in an interview.
Although their business model is still unproven, the founders of AI-Native said their approach helps them stay beautiful.
At Daydream, the fashion shopping startup still in Beta, the employee count is at a low point. “If you want to create this company three to four years ago, you’ll need more people,” said Daydream co-founder Dan Cary, who worked at Google for more than a decade before becoming an entrepreneur.
Or hire Lex as an AI-driven writing assistant Lex with professional writers and academia. It has three employees and its founders who want to keep this way, even if the business grows. “I like to keep the idea smaller over time so that the core groups feel real consistency and real motivation and be able to communicate about the hard things,” said founder Nathan Baschez. “You want the right people, not more.”
It’s not just the raw computing power of AI, keeping the number of employees low – AI’s crash work often spans multiple functions and is done by one person (if any).
In daydreams, even product managers who are not highly skilled can use AI coding assistants to build their own prototypes without having to wait for designers or engineers to build something for them. “It removes the technical barriers to creation” and speeds up the work,” Cary said.
Daydreams do not test one idea at a time like traditional companies, but test 15 to 20 ideas at the same time. “These AI tools are excellent to help you determine if you are making mistakes faster and help you get it right faster,” he said.
With the rise of AI agents, it has become easier to be able to handle multi-step tasks. While the generated AI chatbot may be able to make emails for customers, a proxy can be built to actually send emails, record any replies, and even schedule service calls if needed.
At Sailplane, AI engineer, co-founder and CEO Sam Ramji’s scheduled stealth mode startup construction agent is handled entirely by AI agents, who mimics the tone and rhythm of a professional assistant. Ramji said the company plans to build an agent for any part of the business that doesn’t have ready-made AI solutions.
Sailplane has a group of six people, with the goal of ending no more than 10 people a year.
The founders say that with small sums, they can spend more time building and less time on traditional management. Lex’s Baschez estimates that he spends 70% of his time on engineering work, and he suspects that he will lose the luxury if the business adds more people.
He believes that big companies are difficult to manage and compare them to animals raised on factory farms. “They develop faster than expected, which can lead to various problems,” he said.
AI local companies have the opportunity to keep their organizational charts smaller and even smoother than other startups. While it’s not uncommon for early-stage company employees to wear many hats, one of the promises of AI is that it will expand everyone’s reach, just like the product managers who are now making prototypes on Daydream.
Microsoft’s Spataro said he noticed side companies tend to hire more generalists. He said that because AI tools can often handle narrow tasks traditionally given to experts, “understanding the endpoint of a business proves more valuable.”
However, working at an AI-NENATIC startup still requires at least one special skill: the ability to use AI.
Founder and CEO Andrew Lee said Shortwave is a startup that creates AI-powered to manage emails, cutting its employees from 15 to 15 after being separated from the employees used by AI. Although he plans to add a few more people, keeping the number of people small is not worried about him. He bets on execution quickly to remain competitive – powered by agents rather than mid-level management.
“We need to optimize our team for a few highly cross-functional people, and we can move super quickly,” he said. “Our moat will become speed.”
More stories like this are available Bloomberg.com
