A report, National Machine Intelligence Strategy for the United States, released in March 2018 by the Center for Strategic and International Studies and underwritten by Booz Allen Hamilton makes the case for a comprehensive U.S. framework for maintaining leadership while ensuring responsible development of machine intelligence. We analysed it and prepared a brief summary of the most interesting thoughts and insights.
The United States’ current economic and technical lead in MI uniquely positions us to ensure the technology’s impact will be broadly beneficial, for Americans and the world. Other countries have already established national strategies to capture the benefits of MI and catch up with the United States.
Capturing this opportunity requires we invest in remaining at the leading edge of MI’s development, and plan for the broader changes it implies. A national strategy for MI will provide the United States a platform for creating the policies and investments that ensure MI’s progress is consistent with American goals, norms, and values.
Invest in R&D
Though the United States benefits from strong private- sector investment in MI R&D, companies alone cannot carry the burden of keeping the United States at the forefront of MI innovation. The government has a unique role to play in funding research into defense-specific innovations, systems of MI ethics and control, and high-risk, high-reward long-term research with uncertain commercial returns.
Develop a workforce for machine learning age
USA must adapt to shifting economic circumstances to prepare our workforce for the jobs of the future. This will require a renewed emphasis on computer science education and technical skills to build and maintain MI systems, as well as basic digital literacy and the liberal arts as the economy faces a growing demand for adaptable workers with so skills capable of complementing the operations of MI systems.
Create an open data ecosystem
An essential component for the development of MI systems is access to large amounts of high-quality training data. By expanding government open data initiatives, working with the private sector to support private-private sharing in ways that protect privacy, and supporting the development and deployment of new standards to improve data quality, the government can help MI developers access the data they need to create new and more powerful MI applications.
Create thoughtful public policy
Many U.S. enterprises are held back from capturing the full benefits of MI due to uncertainty over legal and regulatory requirements and a lack of supporting IT infrastructure. To help grow the market for U.S. MI developers and improve productivity throughout the economy, policymakers should consult with industry partners to identify ways to remove legal and regulatory uncertainty, and devise ways of supporting IT modernization within the private sector.
Develop proactive strategies
While the apocalyptic warnings voiced by some in the tech community are overblown, MI systems will raise new challenges in the areas of privacy, algorithmic bias, system safety and control. The U.S. government can help confront these risks by leading in the development of safety, ethics, and control standards for MI, and working with the private sector to develop methods of testing and certification for MI systems.
This strategy should have two overarching goals. The first is to promote the safe and responsible development of MI technology by funding long-term R&D where the private sector is not incentivized to invest, developing a workforce for the MI age, creating dynamic commercial markets for MI technologies to capture the innovation of the private sector, and proactively managing the risks and disruptions that MI will bring.
The second goal is to maintain U.S. leadership in MI by reinforcing our innovation base and establishing strategic partnerships to leverage the comparative advantages of our allies and lead the development of global MI governance.
Author: AI For Security
If you like this article, please subscribe to our monthly newsletter: