AGI, ASI Timelines: Expert Forecasts Show Widening Divergence

Divergent expert timelines for Artificial General Intelligence (AGI) and Artificial Superintelligence (ASI) highlight significant disagreement regarding AI development pace. A recent analysis of 66 forecasts between 2016 and 2023 shows a marked spread in predictions. In 2016, 73% of forecasts predicted AGI by 2040. This figure decreased to 60% by 2023, indicating a shift in consensus.

Conversely, a substantial portion of forecasters now anticipate AGI much sooner. Over a third of experts (35%) in 2023 predict AGI's arrival by 2030. This represents a significant increase from just 9% holding this view in 2016. Furthermore, a smaller but notable segment, 13% of forecasters, expect AGI to emerge before 2027. This suggests a bimodal distribution of expert opinion, with some anticipating rapid progress and others projecting longer development cycles.

The forecasts for Artificial Superintelligence (ASI) exhibit an even broader range. While some experts project ASI within decades, others extend this timeline to centuries. This expanded scope of predictions underscores the uncertainty surrounding advanced AI capabilities. The observed divergence in timelines may stem from differing interpretations of current AI capabilities, varying methodologies for forecasting, or disparate views on critical bottlenecks. These disparate outlooks complicate efforts to prepare for potential societal and economic shifts driven by advanced AI.

The increasing spread in expert opinions raises questions about the predictive power of current forecasting models. Do these diverging timelines reflect inherent uncertainties in nascent technological fields, or do they indicate fundamental disagreements on the definitional criteria for AGI and ASI?