ABSTRACT

This paper studies the economic implications of automation. We consider that automation is affected by disruptive technologies which entail a structural change consisting in the introduction of a new physical capital input (a combination of artificial intelligence and autonomous robots), additional to ‘traditional’ capital assets and labor. This new ‘automatic’ physical capital is assumed to carry out production activities without the need to be combined with labor. We propose a simple production function combining both traditional and new technologies, and show that the consequences of automation depend on the combination of the automatic capital adoption rate and the elasticity of substitution between traditional and automatic technology. We find out that, if the adoption rate is below a threshold that matches the marginal productivity of automatic capital, little effects are derived from automation, independently of the elasticity of substitution. However, if the automatic capital adoption rate is above the threshold level and the elasticity of substitution is higher enough, the automation process can lead to a robocalypse scenario with a total shift of both traditional capital and labor. We estimate, through the benchmark calibration of the model, that the adoption rate threshold for automatic capital is about 22.5%.

摘要

本文研究了自动化的经济影响。我们认为自动化受到破坏性技术的影响，这些技术需要结构性变化，包括引入新的物质资本投入（人工智能和自主机器人的组合），除了“传统”资本资产和劳动力。假设这种新的“自动”物质资本无需与劳动力结合即可开展生产活动。我们提出了一个结合传统技术和新技术的简单生产函数，并表明自动化的结果取决于自动化资本采用率和传统技术与自动化技术之间的替代弹性的组合。我们发现，如果采用率低于与自动资本的边际生产率相匹配的阈值，那么自动化几乎不会产生影响，而与替代弹性无关。然而，如果自动资本采用率高于阈值水平且替代弹性足够高，则自动化过程可能导致传统资本和劳动力发生全面转移的机器灾难场景。我们通过模型的基准校准测算，自动资本的采纳率门槛约为22.5%。自动化过程可能导致机器人灾难场景，传统资本和劳动力都将发生全面转移。我们通过模型的基准校准测算，自动资本的采纳率门槛约为22.5%。自动化过程可能导致机器人灾难场景，传统资本和劳动力都将发生全面转移。我们通过模型的基准校准测算，自动资本的采纳率门槛约为22.5%。