SUSTAINABILITY
TESTING LLMS
it to strategic priorities and establish clear accountability for decisions and outcomes.”
However, a balance is required between those on the ground making operational decisions and executives.
“ The people closest to operational decisions should be involved from the beginning,” she asserts.“ The challenge is not only whether the technology works, but whether employees trust it, understand its limitations and know when to use or challenge its recommendations.”
Vanessa is clear that this requires practical training, transparent communication and evidence from real use cases:“ When people see tangible value in their own work, and responsibility for the final decision remains clear, adoption becomes much more likely.”
Illustrating how AI can be used in conjunction with sustainable practices, at Harvard Business School( HBS), Robert G. Eccles and Shivaram Rajgopal put four widely available LLMs to work on ExxonMobil’ s public disclosures. The authors say the goal was not to produce another ESG score – and it was not to single out ExxonMobil. Robert and Shivaram argue in a blog on HBS’ s website that it was to test whether AI could do something sustainability analysis has long struggled to do at scale: take the environmental and social issues a company itself discloses as financially relevant, map them to specific incomestatement, balance-sheet, cash-flow line items and estimate how strong or weak performance on each would affect the company’ s value. One of the authors of the report had done this kind of analysis by hand before. It took about 100 hours. With AI, the same core work took roughly an hour and parts of it were done in minutes. aimagazine. com 71