ty in order to make it effectively support the mission
engineering processes involved. The initial solution
was a nonteaming AI system. The first attempt to fix
it, by allowing pinning, was a surface, interface-level
solution. The eventual system required deeper teaming in both the AI algorithms and the interfaces,
yielding a highly functional tool that actually
enhanced activity planning productivity.
While AI continues to demonstrate remarkable
achievements, the future lies in its ability to work
well with people. Highly publicized AI examples such
as Watson showcase individual competence, but the
future of such systems lies in their teaming compe-
tence. Jonas Nwuke, an IBM Watson ecosystem man-
ager, emphasizes, “Our perspective is that you can’t
take the humanity out of it. There are a ton of oppor-
tunities that can be met by [technology]. There are a
ton of challenges that can be overcome by it. But, at
the end of the day it is that partnership between man
and machine that matters most.” 9 This sentiment is
consistent with our case for teaming intelligence.
Our recommended pathway forward is for AI to pur-
sue areas that emphasize teaming competencies.
Such an approach would not require a complete redi-
rection of current AI work, but an adjustment and
broadening of that work to include how the new AI
capability will team with people. If done well, the
result should be that the combination of AI and peo-
ple exceeds the performance of either alone.
Current intelligent systems technologies are fundamentally different from human intelligence, and,
more importantly, there is no reason to believe they
are on a convergent path with human intelligence in
the longer term either. This diversity has the potential to provide strength and resilience — but only if
machine and human can work together effectively.
Designing for such mutual cooperation suggests a
radical shift from the traditional divide-and-conquer
approach based on the allocation of function to a
more sophisticated strategy based on supplementa-tion or enhancement, instead of replacement — in
other words, teaming. Interdependence is the essence
of teamwork. AI will only become an effective team
player if it has an understanding of interdependence
and is designed to support management of interdependencies with people.
One of the key challenges AI faces with respect to
teaming is that AI and teaming are often seen as
opposites. This is not the case and in fact develop-
ment of increasingly sophisticated AI capabilities
must go hand-in-hand with increasingly sophisticat-
ed human-machine interaction. Intelligent systems
must be designed from the outset to team with
human capabilities, providing assistance where
human intelligence has limits and leveraging that
intelligence where it is uniquely powerful. Instead of
viewing AI and teaming as opposites, we should view
them as complements, always remembering that no
AI is an island.
We thank Pat Hayes, Ken Ford, Peter Pirolli, Daniel
Duran, Micael Vignati, and Dorrit Billman for
insightful reviews of this work.
qtr_ 2_07/ article_03_2.html.
Adams, S., Arel, I.; Bach, J.; Coop, R; Furlan, R.; Goertzel, B.;
Storrs Hall, J.; Samsonovich, A.; Scheutz, M.; schlesinger,
M.; Shapiro, S. C.; Sowa, J. 2012. Mapping the Landscape of
Human-Level Artificial General Intelligence. AI Magazine
33( 1): 25–42. doi.org/10.1609/aimag.v33i1.2322
Baker, D. P.; Day R.; and Salas, E. 2006. Teamwork as an
Essential Component of High-Reliability Organizations.
Health Services Research 41( 4 pt 2): 1576–98. London: Blackwell Publishing. doi.org/10.1111/j.1475-6773.2006.00566. x.
Billings, C. E. 1997. Aviation Automation: The Search for a
Human-Centered Approach. Mahwah, NJ: Lawrence Erlbaum
Blackhurst, J. L.; Gresham, J. S.; and Stone, M. O. 2011. The
Autonomy Paradox. Armed Forces Journal, October 1, 2011.
Bradshaw, J. M.; Hoffman, R. R.; Johnson, M.; and Woods,
D. D. 2013. The Seven Deadly Myths of Autonomous Systems. IEEE Intelligent Systems 28( 3): 54–61. doi.org/10.1109/
Brooks, Rodney. 2017. The Seven Deadly Sins of AI Predictions. MIT Technology Review, October 6, 2017. www.tech-
Cassidy, Mike. 2014. Centaur Chess Shows Power of Teaming Human and Machine. Huffington Post, December 30,