Startups often fail by making a slightly better version of an incumbent's product. This is a losing strategy because the incumbent can easily adapt. The key is to build something so fundamentally different in structure that competitors have a very hard time copying it, ensuring a durable advantage.

In an era of infinite replicability, startups have two viable paths. They can either operate in stealth with a non-obvious, defensible insight ('a secret incantation'), or tackle an obvious problem and win by completely owning the public narrative. The middle ground is no longer viable.

The pace of AI-driven innovation has accelerated so dramatically that marginal improvements are quickly rendered obsolete. Founders must pursue ideas that offer an order-of-magnitude change to their industry, as anything less will be overtaken by the next wave of technology.

AI-native startups hold a key long-term advantage over established players. Incumbents often struggle to integrate transformative AI because it threatens to cannibalize their existing, profitable business models. AI-native companies, built from the ground up, face no such constraints and can pursue more disruptive strategies.

The core conflict is whether a startup can achieve mass distribution before the incumbent can replicate its core innovation. Historically, incumbents have an advantage because they eventually catch up on technology. AI may accelerate this, making a startup's unique and rapid path to acquiring customers more critical than ever.

To avoid being crushed by AI platform advancements, startups shouldn't compete directly with core models ('under the rock'). Instead, they should find a specific, underserved problem on the outer edge of what's newly possible, where deep user familiarity provides a defensible moat.

Unlike past tech cycles where startups primarily fought other startups (e.g., Facebook vs. Snapchat), today's AI innovators also compete directly with the immense resources, talent, and data moats of established giants like Google and Microsoft.

The AI landscape presents a uniquely challenging competitive environment. While generative AI makes it easier than ever to build and launch products (no barriers to entry), it also eliminates traditional moats like proprietary technology. This forces companies into a state of constant pivoting and feature replication to survive.

Despite the dominance of large AI labs, they face constraints in compute, talent, and focus. Startups can thrive by building highly specialized products for verticals the big players deem too niche. This focused approach allows them to build better interfaces and achieve deeper market penetration where giants won't prioritize competing.