Show Notes
I was pleased to have Colin Carroll join me on the Acquisition Talk podcast to discuss the acquisition of machine learning in the Department of Defense. He is the Director of Government Relations at Applied Intuition, a company that enables autonomous vehicles through simulation development and validation. Before that, Colin had a number of positions including Chief Operating Officer at the JAIC, mission Integration Lead for Project Maven, and 10 years of active service in the Marine Corps.
2:30 -Project Maven started with Bob Work and 10 slides 6:30 - Everyone in the Pentagon's in the fight 10:30 - There's not yet an urgency like in 2009 with MRAP 12:30 - How JAIC operations differed from Project Maven 15:00 - DoD autonomy programs often have zero data 17:00 - How to structure AI/ML programs in DoD 19:00 - The Joint Common Foundations is no more 24:40 - Most DoD's data is owned by industry 27:00 - DoD is buying brittle AI/ML models 29:00 - Competing with GOTS software 31:00 - Separating HW acquisition from SW 37:00 - DoD's $2B AI/ML spending estimate likely high 42:00 - We don't win by reforming SBIR 59:20 - The buzzword of JADC2 1:05:16 - The idea behind Title 10 failed 1:09:50 - Force Design 2030 and the future fight 1:20:10 - How to build a defense team at a tech company
This podcast was produced by Eric Lofgren. You can follow me on Twitter @AcqTalk and find more information at https://AcquisitionTalk.com
More from Acquisition Talk
In this final episode of Programmed to Fail, we explore the true importance of reforming the Planning, Programming, Budgeting, and Execution process in the Department of Defense. Some may say that it is only a poor craftsman who blames his tools, that it is deficiencies in the workforce rather than problems of the acquisition and budgeting systems that are holding weapons innovation back. But certainly, it is not the lack of quality and drive in the people that has held North Korea back relative to their neighbors in South Korea, or that has stymied the growth of nations in the former Soviet Union. It was the ideologies of the political economy thrust upon the people that so devastated their culture.
In this episode of Programmed to Fail, we explore how accounting costs do not reveal the value being generated in the production process and cannot be used, on their own, as a guide for specific choices. Cost is not an objective reality, particularly to those who know the vagaries of cost accounting. Instead, our view of cost depends on subjective use value and is related to the term opportunity cost, or the next-best choice foregone. This chapter reveals that for defense acquisition to truly understand weapons value and leverage the power of commercial markets, it needs to shift away from its obsession with financial metrics.
In this episode of the Acquisition Talk podcast, Dorothy Engelhart joins me to discuss unmanned surface and underwater vessel development in the United States Navy. Dorothy is the Director of Unmanned Systems in the Deputy Assistant Secretary of the Navy for ships, and she has been in this role since 2015. Before that, she was a senior acquisition manager for Marine Corps MDAPs, and had over 20 years of experience in NAVAIR as well as experience on the Hill.
In this episode of Programmed to Fail, we discuss the integral role of rivalrous competition in the discovery of knowledge and the growth of technology. It finds that policy maker's dreams about single best plans and pre-coordinating service behavior to avoid duplication, competition, and overlap is a false economy, one that stamps out the true creative potential of the American people and harms national security.
Welcome to a special series on the acquisition talk podcast that gives you an audiobook tour of my research project titled, Programmed to Fail: The Rise of Central Planning in Defense Acquisition 1945 to 1975. I’m Eric Lofgren of the Baroni Center for Government Contracting at George Mason University. You can find this book for free and over 1,300 blog posts on my website, https://AcquisitionTalk.com.