This editorial appeared in the May 8th, 2025, issue of the Topline newsletter.
Want all of the latest go-to-market insights without the wait? Subscribe to Topline and get the full newsletter delivered to your inbox every Thursday.
If someone says "I love you," are you sure you know what they mean? I used to think I did, but boy was I wrong.
In my first year of university, I took a class called mass media and communication. Our professor made us all write down the meaning of love, then had each of the sixty-five students stand up and read their definition aloud. As definition after definition rolled through the lecture hall, a powerful realization washed over us — not a single person had written the same thing.
Sixty-five people. Sixty-five completely different definitions of perhaps the most universal word in human history. His lesson struck deep: when we communicate, we assume others grasp our exact meaning, but this shared understanding is largely an illusion. The gap between what we say and what others hear is wider than we imagine, breeding misunderstandings that can derail even our most important relationships.
Today, I can't escape the term "AI-First." It's emblazoned across pitch decks, investor memos, and company missions like a secret password to an exclusive club everyone's desperate to join. These companies command premium valuations, attract elite talent with magnetic force, and glide through sales cycles with uncanny ease. It's the golden ticket in tech — flash it, and doors swing open.
But what exactly does it mean to be "AI-First"? This is where certainty crumbles. The definitions splinter and contradict like those 65 versions of love from my college days. Everyone wields the term with absolute confidence, yet beneath the surface lies a fundamental disconnect.
This matters more than we think because speaking a common language isn't just nice — it's the foundation that enables true alignment in business and prevents costly misunderstandings. A term this influential deserves a clear definition, not a fuzzy consensus. So today, I'm going to attempt to pin down what "AI-First" actually means.
A Rorschach Test
Before we can define what being "AI-First" really means, let's examine the dizzying array of definitions already in circulation.
The term gained mainstream traction in 2016 when Sundar Pichai declared Google was shifting "from a mobile-first to an AI-first world" — meaning they would integrate AI into every product experience. Seven years later, Satya Nadella echoed this by "rapidly evolving Microsoft 365 into an AI-first platform." For these tech giants, "AI-First" means comprehensive product integration. But what if these integrations don't actually deliver meaningful results?
Meanwhile, McKinsey takes a different angle, defining it as elevating AI to "the ultimate strategic priority, above everything else." But this approach again lacks a crucial element: results. When AI becomes the goal rather than the means, companies risk chasing technology for technology's sake instead of driving business outcomes.
Then there's Bessemer Venture Partners, who in their "Six Imperatives for AI-First Companies" essay, reserve the term exclusively for companies "in the business of advancing AI as a science," relegating everyone else to merely "AI-enabled" status.
This definition creates more problems than it solves. Companies advancing AI science already have clear labels: "AI research labs," "foundation model providers," and "frontier model developers." They hardly need another designation.
The confusion isn't merely academic. The term now appears everywhere from corporate missions to fundraising announcements, job descriptions to IPO filings. It has become a Rorschach test where everyone sees what they want to see — creating the perfect conditions for expensive misunderstandings and strategic misalignments.
What we need is a definition that is outcome-based, measurable, and universal across industries. Let’s try to find it.
So What Is It?
As we've already established, merely integrating AI features or declaring it your top strategic priority isn't enough. If these efforts don't drive measurable results, you're just AI-washing — slapping a trendy label on business as usual.
"AI-First" also can't simply mean companies founded after ChatGPT's release (that would exclude pioneers like Bolt which was founded in 2022), nor is it companies "whose products significantly improve as LLM models improve" (as one VC suggested in a closed-door session) — virtually any software with AI integration does that.
So what makes a company truly "AI-First"?
It's about transformative throughput — using AI to fundamentally change what your organization can accomplish with the same resources.
We're now in the operationalization phase of AI, where productivity gains aren't incremental — they're revolutionary. Take coding: building an HTTP server once took 160.9 minutes; with AI assistance, it's just 71.2 minutes. Developers using these tools move 50-60% faster. Cursor alone facilitates 1 billion lines of code daily, with 40% generated by AI.
Companies embracing these capabilities ship products at a pace that would have been impossible just two years ago, with engineering teams half the size their competitors require. They've transcended the old constraints of productivity and gained a real competitive advantage because of it.
That's what makes a company truly "AI-First" — not when it was founded or what it claims in investor decks, but the measurable transformation in what it can deliver. This definition is outcome-based, measurable, and universal across industries.
Critics might say this focuses too much on efficiency rather than innovation. But unprecedented throughput enables unprecedented innovation by accelerating experimentation cycles and freeing resources for creative work.
The good news is that this definition democratizes the opportunity. Any company can become “AI-First” - not just AI labs with billion-dollar R&D budgets. The proof isn't words in your pitch deck — it's in your output, in the measurable delta between what was possible before and what you deliver now.
So it's no longer about merely adopting AI — it's about harnessing it to fundamentally transform what your organization can accomplish. Five years from now, companies delivering twice the output with half the resources won't just have a competitive edge — they'll own the market. The rest won't be disrupted; they'll be forgotten.
