Intelligence from frontier models is now a commodity. The real value comes from Forward Deployed Engineers (FDEs) who customize and apply this general intelligence to a company's specific, unique workflows, creating a competitive edge through superior deployment.

Despite proven cost efficiencies from deploying fine-tuned AI models, companies report the primary barrier to adoption is human, not technical. The core challenge is overcoming employee inertia and successfully integrating new tools into existing workflows—a classic change management problem.

The conventional wisdom that enterprises are blocked by a lack of clean, accessible data is wrong. The true bottleneck is people and change management. Scrappy teams can derive significant value from existing, imperfect internal and public data; the real challenge is organizational inertia and process redesign.

Despite powerful models, OpenAI is hiring thousands for roles like 'technical ambassadorship' because enterprises struggle to implement AI. This 'capabilities overhang' shows the biggest challenge isn't model intelligence, but applying it at scale in real-world workflows, which requires significant human support.

Most companies are stuck on providing GenAI licenses and personalized training, which require zero IT involvement. While data and reliable agents are technical hurdles, massive productivity gains are achievable today by solving these simpler, more accessible cultural and educational challenges first.

The primary bottleneck for successful AI implementation in large companies is not access to technology but a critical skills gap. Enterprises are equipping their existing, often unqualified, workforce with sophisticated AI tools—akin to giving a race car to an amateur driver. This mismatch prevents them from realizing AI's full potential.

Despite powerful new models, enterprises struggle to integrate them. OpenAI is hiring hundreds of 'forward-deployed engineers' to help corporations customize models and automate tasks. This highlights that human expertise is still critical for unlocking the business value of advanced AI, creating a new wave of high-skill jobs.

Off-the-shelf AI models can only go so far. The true bottleneck for enterprise adoption is "digitizing judgment"—capturing the unique, context-specific expertise of employees within that company. A document's meaning can change entirely from one company to another, requiring internal labeling.

The strategy to embed thousands of engineers to drive AI adoption is flawed because the necessary talent is scarce. Even top tech companies lack deep benches of expert field engineers capable of solving complex, novel AI problems, making it nearly impossible to scale a services-heavy model effectively.