Even within OpenAI, a stark performance gap exists. Engineers who avoid using agentic AI for coding are reportedly 10x less productive across metrics like code volume, commits, and business impact. This creates significant challenges for performance management and HR.

The cost of re-validating, QA-ing, and re-training internal apps built on a specific LLM far outweighs potential token savings. Once an application is "dialed in" on a model like Claude Opus, the business has little incentive to switch, creating a durable competitive advantage.

Don't view AI through a cost-cutting lens. If AI makes a single software developer 10x more productive—generating $5M in value instead of $500k—the rational business decision is to hire more developers to scale that value creation, not fewer.

As AI token costs become a significant line item, companies will shift from headcount-based budgets to dollar-based budgets. This will force managers to trade B-player employees in roles like QA or customer success to fund unlimited token access for their A-player engineers.

Despite significant history and memory built up in platforms like ChatGPT, power users quickly abandon them for models like Claude or Manus that provide superior results. This indicates that output quality is the primary driver of adoption, and existing "memory" is not a strong enough moat to retain users.

AI acts as a force multiplier for a company's best and most ambitious people, not a tool to make weak performers competent. It allows top talent to automate mundane work and focus on high-value strategy, effectively widening the performance gap between the most and least productive employees.

The productivity gains from individual AI use will become so significant that a wide performance gap will emerge in the workplace. The most talented employees will become hyper-productive and will refuse to work for organizations that don't support these new workflows, leading to a significant talent drain.

Your most skilled AI professionals are also the most mobile. They recognize when their sophisticated work isn't creating value and will leave out of frustration. This turns a project-scaling issue into a critical talent retention problem, as your best people notice when intelligent work goes nowhere.

Data on AI tool adoption among engineers is conflicting. One A/B test showed that the highest-performing senior engineers gained the biggest productivity boost. However, other companies report that opinionated senior engineers are the most resistant to using AI tools, viewing their output as subpar.