The AI revolution fundamentally accelerates innovation, transforming what was an 18-month doubling of capabilities into a six-month doubling due to breakthroughs like deep neural networks and AI accelerators.

The scenario includes a month-by-month analysis of AI developments, starting with agency training and coding improvements, leading to an intelligence explosion where AIs begin to assist in AI research.

AI is going to be the biggest technology shift of our lives, with the potential for explosive growth in the field, unlike previous tech revolutions that relied on hardware distribution.

The exponential increase in computational power is driving advancements in AI, leading to significant improvements in technology and capabilities, and the law of accelerating returns indicates that technological progress, including AI, continues to advance exponentially, as evidenced by a chart I found 45 years ago.

As we approach mid-2027, the key factors in AI research will be the taste of AI agents and the amount of compute available for experiments, as these will determine the effectiveness of automated AI research.

The year 2025 is anticipated to be significant for AI agents, with a focus on their breakout potential in various applications.

AI 2027 is our scenario trying to forecast the next few years of AI progress, aiming to provide a concrete story that shows the transitional steps leading to AGI in 2027 and potentially superintelligence in 2028.

The pace of AI progress is accelerating rapidly, moving on a scale of quarters rather than years.

The current pace of AI progress has been underestimated, and we believe that the speed of AI diffusion into the world is also accelerating.