In 2026, the gap in AI adoption between companies is enormous. Corporations, due to security requirements, are often limited to weak open-source models, even without coding agents. Startups are fully leveraging Claude Code or Codex. A telling example: Boris Cherny, creator of Claude Code at Anthropic, shared that he hasn't written a single line of code by hand for 2+ months — AI agents do everything. Meanwhile, Anthropic is actively hiring engineers, with over 100 open positions on their website.
A paradox? Not really. The engineer's role has simply changed, and it's important for businesses to understand this:
Writing code is losing its value. The value has shifted from HOW to write code to WHAT to build. Architecture, feature prioritization, quality control — that's what engineers are paid for today.
Role boundaries are blurring. If an engineer prompts AI, talks to clients, and decides what to build — how is that different from a product manager? The new engineer is a PM with deep technical expertise. For businesses, this means: rethink who you hire and what skills you pay for.
AI is a multiplier, not a replacement. Anthropic with 3,000 employees is already comparable in impact to Oracle with 162,000. AI doesn't remove people from the equation — it makes each of them exponentially more powerful. The question isn't whether to hire people, but how well your people are equipped with the right tools.
When anyone can make anything — the key differentiator becomes WHAT you choose to make. Taste and strategic thinking become more important than execution skills.
— Paul Graham
For founders, the takeaway is simple: competitive advantage right now isn't about team size — it's about how quickly you integrate AI agents into your processes. Those who do it first will get the same multiplier that gives Anthropic an edge over companies 50 times larger.