Despite rapid technological change since 1971, productivity growth has been at historic lows. Marc Andreessen argues this isn't a technology failure but a policy choice, citing a massive increase in regulations that stifled progress in areas like nuclear power, transportation, and space, leading to economic stagnation.
The threat to established SaaS companies is not just technological but also psychological. Simply adding AI features to an existing product like Photoshop may not be enough if AI creates entirely new workflows. Survival depends on 'human agency'—bold leadership willing to cannibalize existing products and fundamentally reimagine their business for an AI-centric world.
Open source AI models don't need to become the dominant platform to fundamentally alter the market. Their existence alone acts as a powerful price compressor. Proprietary model providers are forced to lower their prices to match the inference cost of open-source alternatives, squeezing profit margins and shifting value to other parts of the stack.
Echoing Don Valentine's VC wisdom that 'scarcity sparks ingenuity,' US restrictions on advanced chips are compelling Chinese firms to become hyper-efficient at optimizing older hardware. This necessity-driven innovation could allow them to build a more resilient and cost-effective AI ecosystem, posing a long-term competitive threat.
The AI competition is not a simple two-horse race between the US and China. It's a complex 2x2 matrix: US vs. China and Open Source vs. Closed Source. China is aggressively pursuing an open-source strategy, creating a new competitive dynamic that complicates the landscape and challenges the dominance of proprietary US labs.
Science fiction depicted AI as either utopian or dystopian, but missed its most immediate social impact: becoming fodder for memes and humor. Platforms like Maltbook, a social network for AIs, demonstrate this unpredictable creativity. This creates a bizarre feedback loop where future models are trained on humorous, human-AI hybrid content, accelerating emergent behavior.
Despite massive investment in chips (NVIDIA) and models (OpenAI), it is not yet clear where long-term value will concentrate. The entire stack is in flux. Models could be commoditized by open source, chips could face historical commoditization cycles, and new AI-native apps could capture the most value. We are only in the early innings of a 30-year shift.