Generative artificial intelligence (AI) leapt to prominence last year. The widespread usage of popular large language model powered chatbots (like ChatGPT) and image generators (ie Midjourney) sparked excitement, controversy, and huge capital investment. Since then, adoption has been widespread and investment has been significant.
However, an array of people and organisations have leveled criticism at generative AI and its applications. The problems raised with the technology range from it being simply inefficient and unappealing to downright unethical. If the supply chain sector is to make the most of its investment into the technology, it needs to avoid making the mistakes already befalling other sectors, where generative AI is actively eroding value—usually for a high price tag.
Generative AI’s big year
Funding for generative AI quadrupled in 2023, and as of February 36 generative AI startups had attained unicorn status. Investment in generative AI startups skyrocketed, from $4.3 billion in 2022 to $21.8 billion last year.
Generative AI’s ability to create (the appearance of) new content, such as numerical data, images, text, audio or video has generated a great deal of investment, excitement, and media attention (in addition to a truly shocking amount of pornography). However, finding ways for the technology to make the leap from curiosity to useful (and, more importantly, profitable) business tool is still an ongoing search.
Clickbait, waffle, and 24/7 content farms
Several companies are providing generative AI tools as a way to supposedly enhance the experience they provide. For example, Ebay has started giving the option for sellers to use AI to automatically generate item listings. However, users have criticised the service for surrounding basic information with overly flowery, poorly phrased “waffle.”
Similarly, AI leveraged to churn out news articles and blog posts as part of a new wave of automated content farms has also faced criticism for flooding the internet with “low quality” articles and “clickbait.” The problem is escalating rapidly, as well, with a recent study conducted by researchers at the Amazon Web Services (AWS) AI Lab finding that a “shocking amount of the web” is already made up of poor-quality AI-generated and translated content.
In short, critics of the technology believe generative AI fails to bring any real value to the areas where it is being deployed. The fact that 40% of supply chain organisations are already investing in generative AI begs the question: what are they planning to use it for? Will it add value to the business?
More pertinently, are there applications for generative AI that actually can add value to the business? Or, is this tech adoption for its own sake going to hurt the organisations that embrace it like it hurt all those kids who wanted a nice weekend out at a Willy Wonka themed experience in Glasgow?
What can generative AI actually do for the supply chain?
The main issue with the more widely known generative AI platforms like ChatGPT is that their outputs are only as good as the data used to train them. Most chatbot AIs currently available to the public are generalists, trained on huge amounts of (stolen) data.
However, if trained on the right, thoroughly vetted data, generative AI can be a useful tool for analysing large, unstructured sets of information. It can rapidly classify and categorise information based on an array of visual, numerical or textual data formats. Then, it can take those large volumes of data and summarise them, extracting key insights and trends. The technology could also potentially assist in quickly pulling relevant information from those datasets in order to provide instant responses by voice or text, which might be useful in allowing workers with a lower level of technical skill to perform higher level tasks.
It can also quickly analyse and modify strategies, plans and resource allocations based on real-time data—much faster, with a much broader pool of information than a human.
Generative AI could also automatically generate content in various forms that enables supply chain managers to automate vendor negotiations according to a preexisting script and set of parameters.
However, it all depends on the quality of the model being used and the quality of the data. Without adequate oversight, direction, and scrutiny, generative AI will erode more value from the supply chain than it creates.
- AI in Supply Chain
- Digital Supply Chain