Companies that started from 2023 onwards are the AI-native generation. Whether your company is younger or older than that, you need to figure out how to become AI-native, or prepare to get outcompeted. But what does an AI-native organization even look like? It’s not just a normal company with chatbots bolted on. The answer is pretty radical: an AI native organisation is effectively an AI multi-agent system.

The office of the CEO should be AI. The engineering work should be performed by AI. Everything in between should be AI. The humans in the company are still there, but only to extend, maintain and govern the AI, or as the CEO of Zapier Wade Foster put it, to “build the robot.”

From Specialists to Generalists

The days of the specialist are numbered. For information work at least, AI is already faster, cheaper, better informed and nearly as smart. And getting smarter every other week. What humans have to contribute isn’t deep specialization, but adaptability, fast learning and broad cross-domain ability. Every human left in the organisation is a manager, strategist and architect (of agents).

McKinsey popularized the idea of the “T-shaped” skills profile, a person with broad cross-domain knowledge and a deep specialization that is accretive to the organization. With AI, specialization is nice but being a competent generalist is the new baseline. Perhaps McKinsey will soon declare the need for an “em-dash” skills profile!

This shift is going to be psychologically jarring for anyone with decades of hard-won experience, who’s sunk a large chunk of their life into honing a craft. It was jarring for me, and for others I know. But as Sam Altman said “this is gonna be more like the Renaissance than the Industrial Revolution.”

We’re still figuring out AI-native hiring at Jentic, but we’re zeroing in on:

  • Generalists: T-shaped people who are willing to become em-dash-shaped people.
  • Curiousity: High-trajectory, inately curious fast-learners.
  • Agency: Doers, not those living in Dr Seuss’s Waiting Place, waiting for something good to happen to them.

Also, everyone has to be super-smart, but it feels transitory to make that a core requirement when the leading models have already achieved 130+ IQs, and might surpass Einstein before the end of the summer.

The Sovereign Stack

There’s a new sheriff in town: AI ops. This team sits at the crossroads of engineering and business ops, focused on automating everything. They are the ultimate robot-builders, and their mission is to minimize future hiring. They eliminate spreadsheets, cancel SaaS subscriptions, and divert emails to shared inboxes that are triaged by AI. They lurk around trying to notice when human life-hours are spent copy-and-pasting between SaaS applications, email and spreadsheets. In this future, the CTO doesn’t get email spam, because he/she doesn’t have an email address, and OCTO@yourcompany.com is monitored only by AI.

SaaS no longer fits. The AI-native org builds a bespoke, evolving super-app, unified by a single data platform. No more data lakes, no more silos. Internal AI gets access to a complete data monolith, planning and executing without friction.

The three-thirds workday

The working day is no longer about getting work done. Instead it’s about catching up with AI, learning at speed from your peers, and then getting AI to do it.

  • First third: Solo Learning

    Mornings are for catching up on AI news, techniques, and sharpening your own saw. It’s not personal development, it’s survival.

  • Second third: High-Bandwidth Collaboration

    Groups learn faster than individuals. Afternoons are for peer learning, reasoning through 1st order and 2nd order consequences, collaborating on prompt and context management.

  • Final third: High-Leverage Execution

    A few hours to use AI to get your days work done in a fraction of the time.

On Ownership

There’s an elephant in the room when it comes to AI transformation: it’s all about replacing human labor. Yes, productivity should increase, but AI will handle that workload too. On the current trajectory, humans will not upskill past AI. So how can we ask employees to help build an AI native company?

The answer is ownership. The AI native company is built by smart humans who are thinking about the post-labor economy, who understand that asset ownership is more important than wages, and who realize they have an opportunity not just to help build the robot, but to own a share of it.

The Human-AI Partnership

Humans are evolved to empathize with humans, to anticipate needs, to read between the lines. But the performance of our new AI colleagues depends on us being attentive to their needs - to make sure they are set up for success with the right prompts and context. To properly read their work, and give them thoughtful feedback.

It’s not easy to transfer empathy to a machine. We need to either get over ourselves, or dress AI up like a human (an ethically perilous temptation that might transfer moral agency), or to view the AI as a natural extension of ourselves.

Frankly, this is a total inversion of our relationships with computers. What was our tool in 2024 is becoming our colleague in 2025, and might be our boss by 2026. The future belongs to those who are on the path to figuring it out.