The Rise of the Joint-Mission Startup
And why the biggest AI opportunities in the world can’t be solved with SaaS
Stop reading now if you think AI is going to make you rich quickly. There (still) is no such thing as a free lunch.
Enterprise SaaS thinking isn’t going to work for AI.
The enterprise SaaS playbook that was written over the last 20 years is narrow and focused on incremental efficiency gains delivered through scalable point solutions. It is limiting and zero-sum.
AI is positive-sum. Its full potential goes far beyond efficiency gains - it compels us to reimagine how we work today entirely, changing the base unit of work at a company from a team of humans to a mixed team of humans + machines.
A lot of the value of AI will be realized in the enterprise. In fact, the world’s largest enterprises are positioned to capture enormous value. It’s a once in a lifetime opportunity for CEOs at the helm of the world’s largest companies to make bold bets and leapfrog competitors. But CEOs don’t buy SaaS solutions. You don’t go from a human-human org to a human-machine org by making individual workers more efficient. You reimagine how people team in an AI-native universe.
Consultants don’t know the answer. Ask any engineer or AI scientist where their smartest friends are working – it’s not at the CEO’s “go-to” consulting players. The talent lies either at startups or at Nvidia, Meta, Google and Microsoft.
But most startups aren’t tackling this problem and the big tech giants have no incentive to really partner to deliver a moonshot outcome for one amongst the thousands of CEOs and CTOs in their customer base. Startups continue to build enterprise SaaS class solutions to capture quick value from AI-driven efficiency gains. This makes sense - a small team can move quickly and build on top of LLMs to capture ground quickly. Grow fast → Raise capital → Win. But those wins are proving short lived. Fast rise = fast fall in LLM-wrapper SaaS.
Ultimately, the rules have changed for building a generational company. The biggest opportunity is reshaping cross-team workflows. Parker Conrad talks about Compound Startups in this excellent talk. A compound startup is a startup that, “instead of doing one very narrow thing, tries to build a whole set of point solution systems in one coherent product to tackle a much larger problem for a business”. Compound AI startups can become true partners to the enterprise, building AI systems and shaping the org of the future. Shared foundational AI minds will transfer learning across teams, reshaping how teams are built, how they interact and how they work. But it requires patience, investment and a willingness to go against the grain.
To deliver the Apollo program, NASA engaged 20,000 partner firms. Grumman built the LEM (the lunar lander). This wasn’t your vanilla vendor-customer relationship. It was a joint mission. It was positive-sum.
Ambition isn’t exclusive to Silicon Valley. At Arena, we are lucky enough to work with visionary leaders in the Fortune 500 with inspiring, expansive AI ambitions to reimagine the very core of what they do and how they make money – not just LLM productivity boosts for business support functions.
We are going to see more “Apollo” style bets from the captains of Fortune 500 companies. To make these bets, they don’t need another SaaS product - they need a deep AI partner. A true partner, not a vendor. One that looks more like Lockheed Martin’s Skunk Works than BCG or Microsoft. Those guys aren’t going to deliver you an SR-71.
Building AI skunkworks for global enterprises also calls for a different breed of engineer – one part mad scientist, two parts hacker-engineer and one part Indiana Jones. If you want to work on State of the Art AI in joint-missions with some of the world’s most important enterprises and go to places like this for work – give me a call.
