Can large language models help solve the cost problem for the right to explanation?

Area of Science:

• Artificial Intelligence
• Ethics in Technology
• Decision Systems

Background:

• A consensus exists on the moral right to explanations for automated decision systems.
• Providing explanations for high-stakes automated decisions presents a significant cost challenge.
• This challenge, termed the 'cost problem,' impacts organizations using these systems.

Purpose of the Study:

• To investigate the potential of large language models (LLMs) in addressing the cost problem of automated decision explanations.
• To evaluate the feasibility and ethical implications of using LLMs for generating decision explanations.

Main Methods:

• Exploratory analysis of large language models' capabilities in generating explanations.
• Conceptual examination of the ethical trade-offs involved.

Main Results:

• Large language models show promise in potentially overcoming the cost problem associated with explanations.
• The use of LLMs for this purpose may incur serious ethical costs.

Conclusions:

• LLMs present a potential pathway to mitigate the financial burden of providing explanations for automated decisions.
• The ethical implications, including fairness and transparency, require careful consideration and further research.