Manufacturers are implementing planned generative AI initiatives more slowly than expected due to concerns about accuracy, according to a study by Lucidworks released on Wednesday. The study, which surveyed over 2,500 AI technology decision-makers worldwide, found that 58% of manufacturing leaders plan to increase spending on AI in 2024. This is lower than the global average of 63% and the U.S. average of 69%.
In 2023, 93% of all leaders, including those in manufacturing, had planned to boost AI spending.
Significance
PitchBook data from last week highlighted a quarterly increase in U.S. venture capital funding, primarily driven by major investments in AI companies. Investors are banking on AI startups, hoping that the adoption of AI will yield substantial returns. Nearly 50% of manufacturers worldwide reported cost savings this year after adopting AI initiatives, according to Lucidworks’ study.
“While many manufacturers recognize the potential benefits of generative AI, challenges such as response accuracy and cost are causing them to take a more cautious approach,” said Mike Sinoway, CEO of Lucidworks.
Context
Generative AI processes inputs or prompts to produce new content based on past training data. However, it sometimes generates inaccurate or nonsensical outputs known as hallucinations. While 36% of all respondents expressed concerns about response accuracy due to hallucinations, a higher number of manufacturing respondents, 44%, shared this concern.
Despite only 20% of planned AI projects being implemented by manufacturers in the past year, 55% of them feel that they are on par with their peers in AI adoption.
Cost Factor
Over the past year, 70% of manufacturing companies have opted for more expensive commercial AI models. Lucidworks suggests that a shift to more accessible open-source models could occur if they prove to be as efficient and more resourceful at a lower cost.
While manufacturers are optimistic about the cost benefits of AI, they aim to maximize its value amid smaller planned spending this year.
