Despite its potential to streamline R&D, outsourcing general artificial intelligence (AI) in innovation pipelines could be catastrophically counterproductive, writes software engineer Ashish Pawar in VentureBeat.
Gen AI, driven by a vast dataset, is only capable of predicting improvements rather than creating a genuinely new product or concept and is therefore limited in ability to generate game-changing output.
One key danger of a world where AI is the default for R&D is that content is processed in a highly convergent way, leading to template-style homogenization of the market. If every company in a category uses an AI system to design its products, all will generate variations of the basic structure.
The process of generating breakthrough innovations is undermined because AI is not designed to embrace ambiguity or complexity, which flourishes in human researchers' proven ability to learn from unexpected or ambiguous findings.
The more a company becomes reliant on AI, the less capable of innovation it may become, as human skills and engagement disappear from the workplace.
Pawar argues that rather than turning to AI for answers, companies should focus on steering the development potential of AI with human empathy, vision and a long-term commitment to investing in human creativity.