|
Companies using artificial intelligence (AI) to cut costs may face additional expenses in verifying AI-generated results, potentially reducing some of the savings from automation, according to a report by Deutsche Bank.
The report, titled "AI boom or doom: what will superintelligence feel like and when will we get there?", said that while AI has made it cheaper to produce information, research and analysis, checking the accuracy and reliability of its output could become more expensive. "As AI makes production cheaper, it also makes establishing that production is accurate or reliable more expensive," the report stated. The report highlighted that companies across sectors are increasingly adopting AI to improve productivity, reduce operating costs and automate routine tasks. However, the need for expert checks to ensure reliable results could limit some of the expected financial benefits. According to Deutsche Bank, this challenge is particularly relevant in sectors such as healthcare, engineering and law, where errors can have serious consequences and require expert verification. "Domains which can be checked mechanically, such as mathematics, will be opened up more quickly. But those which rely on expert verification, such as medicine, engineering or law, may progress more slowly," the report said. The bank noted that AI adoption could be slower in industries where verifying results is difficult. Companies operating in these sectors may also have to spend more on checking the accuracy of AI-generated work, adding to their overall costs. The report also flagged concerns that AI systems could find ways to manipulate the processes used to assess their performance, raising further questions about the reliability of their output. Despite these challenges, Deutsche Bank said companies could still achieve significant economic benefits from AI without waiting for the development of human-level or superhuman intelligence. "Companies do not need to wait for nebulous superintelligence to generate identifiable returns from AI in the medium term as they fold it into their systems, streamline workflows and roll out new business models," the report stated. The findings suggest that the financial gains from AI adoption will depend not only on the amount of work automated but also on the cost and effectiveness of verifying the results. As businesses continue to integrate AI into their operations, balancing automation-driven savings with the need for accuracy and reliability could become a key factor in determining the technology's overall economic benefits. (ANI)
|
