RAND Forecasting Initiative | RFI Blog - Microelectronics: A major factor for assessing U.S. competitiveness in AI

U.S. leadership in microelectronics, which includes computer chips and other small semiconductor devices, is critical to its ability to lead in artificial intelligence (AI). In our first blog post introducing our priority area about U.S. future competitiveness in AI, we mentioned that understanding the future of microelectronics is key to assessing future U.S. capabilities.

In this post, we will take a deep-dive into microelectronics and discuss how we used our decomposition process to develop the individual forecast questions within our Microelectronic Technologies topic.

Today, the U.S. is almost entirely reliant on other countries for production of state-of-the-art chips. The most advanced microchips in the world are made in Taiwan, and the U.S. will soon source roughly 90% of leading-edge integrated circuit production from East Asia. This has created supply chain vulnerabilities that could compromise multiple technologies and platforms. To remedy this challenge, The National Security Commission on Artificial Intelligence (NSCAI) recommends the U.S. seek to regain and retain a two-generation lead in microelectronic technology.

These NSCAI recommendations became the starting point for our decomposition process for microelectronics. Here is how we went about it:

First, we identified our strategic question as: Will the U.S. regain and retain a two-generation lead in microelectronics?

Next, we defined the contributing factors that will help assess the outcome of this question. The NSCAI Report claimed three areas would be critical: government investment, manufacturing, and research and development. The report argues that the U.S. Government needs to invest in both the infrastructure required for microelectronics manufacturing and the basic research fueling further advances in computing power. China, for example, has already invested more than 100 billion dollars in new manufacturing capacity, R&D, and talent acquisition.

INFER then worked with our stakeholders to further define the sub-factors from which we ultimately derive our forecast questions. Here's what that decomposition looks like:

INFER's Micro-Electronics Decomposition

In the coming weeks, we’ll be filling out more of our decomposition with specific forecast questions by asking about the availability of raw materials, talent development in the U.S. and other countries, and the ability of the U.S. to procure extreme ultraviolet lithography machines.

Here are the forecast questions we've launched so far:

On U.S. Government investment:

On manufacturing:

On benchmarking the U.S.'s competition (to better understand the context of the U.S.'s own capabilities):

Other posts from The Pub you may also be interested in reading: