A new market has emerged where defunct startups sell their entire operational histories—including codebases, internal communications, and go-to-market data—to AI labs and data brokers. This creates a new form of salvage value, turning years of failed effort into a valuable corpus for training next-generation models.

As startups build on commoditized AI platforms like GPT, product differentiation becomes less of a moat. Success now hinges on cracking growth faster than rivals. The new competitive advantages are proprietary data for training models and the deep domain expertise required to find unique growth levers.

The AI revolution may favor incumbents, not just startups. Large companies possess vast, proprietary datasets. If they quickly fine-tune custom LLMs with this data, they can build a formidable competitive moat that an AI startup, starting from scratch, cannot easily replicate.

With public data exhausted, AI companies are seeking proprietary datasets. After being rejected by established firms wary of sharing their 'crown jewels,' these labs are now acquiring the codebases of failed startups for tens of thousands of dollars as a novel source of high-quality training data.

The venture thesis for AI is shifting towards companies that cannot be easily absorbed as features by large platforms like OpenAI. Investors are targeting startups with defensible moats derived from navigating complex regulations (e.g., medical) or owning unique, proprietary datasets that are difficult to replicate.

As AI becomes commoditized, the key differentiator will shift from *if* a company uses AI to *how good* its underlying data is. AI is only as effective as the context it's given, meaning companies with unified customer data will pull far ahead of those without it.

AI startups are achieving unprecedented 10-50x growth by securing massive, eight-figure contracts from major AI labs. These labs have extreme urgency and large, net-new budgets to acquire key technology or data, creating a powerful new sales channel.

Haystack's "Big Token" thesis posits that large AI foundation models (like OpenAI) will acquire startups not for their applications, but for their unique, proprietary data sets ("tokens"). This mirrors the Big Pharma model of buying smaller biotech firms for their R&D and drug assets.

Since all competitors can access public data through common AI tools, it offers no sustainable advantage. To drive more pipeline and revenue, companies must seek out and integrate proprietary or non-public data sources aligned with their Ideal Customer Profile (ICP), creating a unique data asset for their AI to leverage.