据首席财务官德莫特·麦克多诺透露，纽约梅隆银行正在加大力度监控人工智能成本，以避免超支。
Bank of New York Mellon Corp. is increasing efforts to monitor artificial intelligence costs so as to avoid overspending, according to Chief Financial Officer Dermot McDonogh.
麦克多诺在接受彭博电视采访时表示，该银行已开发出工具来追踪成本，并将任务路由至最合适的模型。
The lender has developed tools to track costs and route tasks to the most suitable models, McDonogh said in an interview on Bloomberg Television.
“我们已经变得相当严谨，并开发了一些复杂的遥测手段，以便追踪我们的支出情况和令牌使用情况，”他说。麦克多诺表示，随着成本因素发挥作用，将任务导向合适模型的努力“随着时间的推移将变得更加重要”。“我们不会对廉价查询使用昂贵的模型，”他说。
“We’ve become quite disciplined and developed some sophisticated telemetry so we can track how we’re spending and how we’re using our tokens,” he said. Efforts to direct tasks to appropriate models will also become more important “over time as costs come into play,” McDonogh said. “We don’t use expensive models for cheap queries,” he said.
随着大规模部署模型的成本飙升，全球最大银行如何使用AI正受到审视。尽管AI被吹捧能改进流程和提高效率，但该技术将在多大程度上影响岗位和工作一直是令人焦虑的来源。
The ways in which the world’s biggest banks are using AI are under scrutiny as costs of deploying models at scale soar. And even though AI is touted to improve processes and efficiency, the extent to which the technology will impact roles and jobs has been a source of angst.
麦克多诺表示，该行的内部AI系统利用超大规模云服务商的模型，且对任何单一供应商保持“中立”。他称，AI仅占该行每年40亿美元技术支出的一部分，该行拥有约400个AI智能体。
McDonogh said the bank’s internal AI system harnesses models from the hyper-scalers and is “agnostic” to any one provider. AI is only a portion of the lender’s $4 billion annual technology spend and the bank has around 400 AI agents, he said.
“我们为正确的项目或正确的计划使用正确的模型，”他说。“我们不对任何单一特定模型保持忠诚。”尽管如此，该行不希望对AI成本过于僵化的关注阻碍使用。
“We use the right models for the right project or the right initiative,” he said. “We don’t have loyalty to any one particular model.”Still, the lender does not want an overly rigid focus on AI costs to stymie usage.
“它将真正为我们未来的许多计划提供动力，”麦克多诺说。“我们不将其与效率或裁员联系在一起讨论。我们将其与技能提升和试图做得更好联系在一起讨论。”
“It’s going to really power a lot of initiatives for us going forward,” McDonogh said. “We don’t talk about it in relation to efficiency or job cuts. We talk about it in relation to upskilling and trying to do better things.”