Early-stage AI startups should resist spending heavily on fine-tuning foundational models. With base models improving so rapidly, the defensible value lies in building the application layer, workflow integrations, and enterprise-grade software that makes the AI useful, allowing the startup to ride the wave of general model improvement.

While it's tempting to build custom AI sales agents, the rapid pace of innovation means any internal solution will likely become obsolete in months. Unless you are a company like Vercel with dedicated engineers passionate about the problem, it's far better to buy an off-the-shelf tool.

The opportunity cost of building custom internal AI can be massive. By the time a multi-million dollar project is complete, off-the-shelf tools like ChatGPT are often far more capable, dynamic, and cost-effective, rendering the custom solution outdated on arrival.

The rise of AI agents introduces a new strategic layer for marketers. They must now decide when to buy out-of-the-box agents, use workflow tools for assembly, or custom-build agents for niche, proprietary tasks. This "build vs. buy" competency is becoming a key marketing differentiator.

Advocates for buying most AI agents off the shelf to leverage existing solutions. Building should be reserved for the small fraction where no suitable tool exists, where you can replace a mediocre incumbent, or where proprietary data is a key advantage.

The enduring moat in the AI stack lies in what is hardest to replicate. Since building foundation models is significantly more difficult than building applications on top of them, the model layer is inherently more defensible and will naturally capture more value over time.

The initial AI rush for every company to build proprietary models is over. The new winning strategy, seen with firms like Adobe, is to leverage existing product distribution by integrating multiple best-in-class third-party models, enabling faster and more powerful user experiences.

According to an MIT report, enterprise AI projects led by external vendors are twice as likely to succeed as those built by internal teams. This is primarily due to a talent gap, as top-tier AI engineers and developers are concentrated in startups, not large corporations.

Forgo building custom AI tools for common problems. Instead, purchase 90% of your AI stack from specialized vendors. Reserve your in-house engineering resources for the critical 10% of tasks that are unique to your business and for which no adequate third-party solution exists.