Renowned investor Elad Gil states that the technological advancements of the past six months may be one of the most significant periods in human history
AI models and agents have introduced recursive self-improvement tools, currently in the very early stages of exponential growth, yet few outside the tech circle are aware.
In market mechanisms, AI capital is accelerating towards models and agent infrastructures with self-iteration capabilities, with funds flowing from application layers to foundational recursive improvement tools. Leading laboratories benefit from the compounding effect of technology, while traditional AI application projects are under pressure due to slower iteration speeds.
Source: Public information
ABAB AI Insight
Elad Gil, as an early investor, has been deeply involved with companies like Twitter and Airbnb. His assessment continues his keen observation of exponential technological inflection points, similar to his past insights during the early phases of mobile internet and cloud computing.
In terms of capital pathways, leading AI laboratories are concentrating resources on recursive self-improvement cycles (model optimizing models, agent improving agents), motivated by the potential for positive feedback loops that could yield compounded growth far exceeding linear progress. Establishing a leading advantage early can create insurmountable barriers.
Similar cases include the acceleration of self-iteration in deep learning from 2012 to 2017, and AlphaGo's transition from supervised learning to self-play; AI is currently at a critical point of shifting from human-led supervision to highly autonomous recursive improvement.
Essentially, this represents a technological substitution: AI development is transitioning from human-led iteration to autonomous recursive improvement by models/agents, with the mechanism being that once the toolchain matures, self-optimization cycles significantly enhance iteration speed, allowing organizations mastering core recursive capabilities to gain exponential capital return advantages and reshape the entire paradigm of technological progress.
ABAB News · Cognitive Laws
During exponential takeoff, the quietest moments are the most dangerous.
The deeper the recursion, the more intense the compounding; the stronger the self-improvement, the more peripheral the human role.
Astute observers see inflection points early, while ordinary people learn history late.