Industrial Use of Generative AI : Opportunities and Risks
Prior to Generative AI , the AI landscape was dominated by customized solutions where machine learning experts would build distinct models for specific tasks . These efforts , though successful in fields like computer vision and predictive maintenance , often led to narrow applications . Each AI implementation was not only constrained by its specificity but was also costly and timeconsuming , limiting its widespread adoption .
FMs marked a significant shift in the world of AI . Instead of training specialized models for each task , FMs are trained on vast amounts of data spanning many different domains and sources , resulting in a versatile and expansive model . While the initial investment in building FMs is high 16 , they serve as a foundational platform upon which a plethora of AI applications can be built . This minimizes -- and in some cases , eliminates -- the need for task-specific training . They have revolutionized the AI landscape , making it more cost-effective , faster and more sophisticated AI deployment across industries .
3 ON OPPORTUNITIES
The potential industry applications of generative AI are vast and multifaceted . From the inception of a product to its eventual disposal , this technology is redefining traditional processes . Its capabilities extend from design and engineering to sourcing and supply , manufacturing , servicing , and more .
In product requirement phase for instance , Generative AI assists in streamlining requirement solicitation process and facilitates a systematic approach towards defining product specifications . It can generate comprehensive questionnaires tailored for requirements gathering , ensuring that all aspects of a product ' s needs are addressed . By customizing questions for specific organizational roles , it ensures both thoroughness and relevance , enhancing and accelerating the preliminary stages of product development .
During the product design phase , Generative AI ’ s ability to churn out numerous design concepts and variations in a short span of time is transformative . Designers input parameters and requirements into a model , and in return , they receive a multitude of design variations . Not only does this expedite the brainstorming phase , but it also allows for the exploration of innovative designs that might not have been considered otherwise 17 . Additionally , it can generate scenarios for validating design options , thus seamlessly bridging the gap between conceptualization and user research .
Generative AI also transforms design validation process . Traditional methods often rely on a set of pre-defined test cases , which might miss out on edge scenarios . Generative AI , with its capacity
16
https :// www . forbes . com / sites / craigsmith / 2023 / 09 / 08 / what-large-models-cost-you--there-is-no-freeai-lunch /? sh = 30c969184af7
17
MIT Technology Review Research . Generative AI : Industrial Design and Engineering . s . l . : MIT Technology Review , 2023 .
Journal of Innovation 5