Alignment of bricolage practice: How supply chain managers mitigate risks through human–AI teaming
Tiancheng Cao, Wenxin Mu, Pei Wu, Juanqiong Gou, Minghong LiuBackground
As organizations embed algorithmic agents in operations, alignment is presumed to make human–AI teams safer and more effective. Yet it leaves the orientation of alignment unexamined, that is, what humans and agents actually coordinate around.
Objective
Drawing on bricolage theory, this study distinguishes three orientations of alignment: tangible-resource, opportunity-oriented, and risk-oriented, then examines how each relates to collaborative fluency and to financial and inventory risk.
Methods
We analyzed 141 simulated supply-chain teams, each pairing a human manager with seven rule-based AI agents between 24 and 47 cycles. Behavioral logs were coded for alignment orientation, while collaborative fluency, risk behavior and outcomes were computed from logs by hierarchical regression.
Results
Tangible-resource alignment was associated with lower inventory redundancy and smaller drawdown. Risk-oriented alignment was associated with greater net-asset volatility and drawdown, associations that weakened once financing was controlled. Opportunity-oriented alignment showed no associations. No orientation explained collaborative fluency, and financing dominated financial risk.
Conclusions
Whether human–AI alignment reduces risk depends on its orientation. Tangible-resource alignment can buffer operational risk, but risk-oriented alignment may mark risk-taking rather than control it. Effective teaming depends less on more alignment than on aligning in right domain, with restraint in financing as the primary safeguard.