Teams rationalize failures by blaming others, creating false internal narratives. Leaders must combat this "storytelling" by seeking unvarnished truth directly from customers and data, bypassing the echo chamber that obscures product-market fit and competitive realities.
Just as developers use various databases for different needs, AI applications will rely on a "constellation" of specialized models. Some tasks will require expensive, high-reasoning models, while others will prioritize low-latency or low-cost models. The market will become heterogeneous, not monolithic.
The idea that AI leads to job cuts misses the competitive dynamic. Since all companies have access to AI, efficiency gains will be reinvested to out-compete rivals, not just pocketed as profit. This escalates competition, turning AI adoption into a strategic imperative for survival and growth.
To prevent a culture of blame, Sierra holds public "lessons learned" sessions for any failure, from lost deals to bugs. This frames failure as a collective responsibility of the team, not an individual's fault. The focus is on fixing the underlying system, fostering paranoia about processes, not people.
The flood of VC money in AI isn't just funding winners; it's creating highly-valued competitors that are too expensive for incumbents to acquire. This is preventing the natural market consolidation seen in past tech cycles, leading to a prolonged period of intense competition.
Many enterprises explore AI due to pressure, not strategy, a phenomenon called "AI tourism." To avoid wasting resources on these tire-kickers, Sierra requires paid proofs-of-concept. The payment, even if modest, serves as a powerful filter for serious buyers with a real intent to deploy.
To credibly sell to the largest enterprises from day one, Sierra intentionally hired experienced executives. The crucial filter was selecting for "competitive intensity" and high agency, avoiding the political mindset often associated with big-company hires. This allowed them to land massive customers early.
Experienced founders have a critical advantage: they can personally vet key hires based on years of observation. First-time founders often rely on their board's recommendations, which can lead to mismatched hires ("organ rejection") because they lack the firsthand context to judge fit.
The hype around future model improvements overshadows a key reality: current models are already "sufficiently intelligent" for countless valuable tasks. Even if all AI innovation stopped today, we could still unlock trillions in economic value just by integrating existing technology across the economy.