ESMT Berlin's new research shows that using machine learning in the workplace can improve the accuracy of human decision-making, but at the same time, it can also lead to increased cognitive effort when making decisions.
These findings come from the research of ESMT Berlin Professor of Management Science Tamer Boyaci and Francis de Véricourt, as well as former ESMT postdoctoral researcher and current Ozyegin University Assistant Professor of Operations Management Caner Canyakmaz. The researchers wanted to study how machine-based predictions affect the decision-making process and outcomes of human decision-makers. Their paper was published in the journal Management Science.
Interestingly, when professionals are under cognitive constraints, such as time pressure or handling multiple tasks simultaneously, using machines increases their workload. However, decision-makers often choose to introduce artificial intelligence to alleviate their workload when faced with high workloads. The research suggests that in such cases, using artificial intelligence may have the opposite effect.
The researchers also found that while machine-generated recommendations can improve the overall accuracy of human decision-making, they can also increase the likelihood of certain types of errors, such as false positives. In this study, machine learning models were used to identify differences in accuracy, bias, and cognitive effort levels between humans and machine-assisted decision-making.
"Recently, many organizations have expressed concerns about the rapid adoption of artificial intelligence technology, fearing that it may eventually replace humans in certain tasks. However, in fact, artificial intelligence can complement human work very well," said Professor Francis de Véricourt.
The researchers state that their research clearly demonstrates the value of collaboration between humans and machines for professionals. However, individuals should also be aware that while machines can provide incredibly accurate information, humans often still need cognitive effort to evaluate their own information and compare the conclusions of machines with their own before making decisions. The researchers note that the required cognitive effort increases when people are under pressure to make decisions.
Professor Boyaci said, "Due to their incredible computational power, machines can perform specific tasks with incredible accuracy, while human decision-makers are flexible and adaptable but limited by their cognitive abilities, so the abilities of machines and humans are complementary. However, humans must be cautious when using machines and know when they are effective and when they are not."
Using doctors and patients as an example, the researchers' findings suggest that using machines will improve the overall accuracy of diagnoses and reduce the number of misdiagnosed patients. However, if the disease has a low incidence rate and time is limited, introducing machines to assist doctors in diagnosis will lead to more misdiagnoses, and the ambiguity caused by machines may increase the workload for humans.
The researchers state that their findings provide hope and warnings for those who want to use machines in their work. The positive aspect is that the average accuracy rate improves, and when machine suggestions tend towards the expected results, the error rate decreases, and individuals become more "efficient" as cognitive effort is reduced.
However, incorporating machine-based predictions into human decision-making is not always beneficial, both in terms of reducing errors and cognitive effort. In fact, introducing machines to improve the decision-making process may have the opposite effect, as it increases certain types of errors and requires more time and cognitive effort for decision-making.
These findings highlight the significant impact of machine-based predictions on human judgment and decision-making. They provide guidance on when and how to consider machine suggestions and offer guidance on human-machine collaboration.
Alphabet Q1 paid subscribers reached 350M, fueled by YouTube & Google One. YouTube ad revenue slightly missed at $9.88B amid Premium growth. Total Q1 revenue hit $109.9B, beating forecasts.
Anthropic releases Claude connectors for Adobe, Blender, and Ableton, enabling AI workflow automation and app control. It also sponsors the Blender Development Foundation.
Apple introduces an App Store model splitting 12-month plans into monthly installments. On iOS 26.4+, it lowers upfront costs without reducing total price, adding billing tracking.
Google launches Aadhaar digital IDs in Wallet for India, enabling secure QR/NFC verification. A phased partner rollout targets key sectors, while passport-based credentials expand globally.
YouTube tests "Ask YouTube," an AI tool blending text and video for complex queries. Initially for US Premium users, it supports interactive follow-ups, expanding Google's AI search.
Google grants US DoD classified AI access, joining OpenAI and xAI in defense deals amid Anthropic's dispute. Contractual safeguards against surveillance and lethal weapons remain unenforceable.
Mistral AI launches Workflows, a Temporal-powered orchestration platform for secure, Python-driven enterprise AI automation featuring human-in-the-loop approvals.