Manhattan Associates, a supplier of supply chain management software solutions and market share leader in warehouse management systems (WMS), held its Momentum user conference last week. Sanjeev Siotta, Manhattan’s chief technology officer, addressed ChatGPT’s role in supply chain management during one of the keynotes on the first day of the conference.
For those of you who haven’t been paying attention to tech news lately, ChatGPT (a type of so-called generative AI) is the most important technology announcement since the Apple iPhone. This is the fastest growing application of all time. In just two months he reached 100 million users. ChatGPT is an AI text generation bot built on the Large Language Model (LLM) family. These models have been trained on massive amounts of data so they can understand text prompts and generate answers to them. However, to say that the application generates the answer is a huge understatement. You can generate an entire article in just seconds.
I played around with a supply chain themed bot. My conclusion is that my job as a writer is not in immediate danger. For certain themes, the engine did not have the correct data to generate useful articles. If the subject was a little broader, the articles that were generated were very well written, but contained a lot of fluff and didn’t delve into what I consider to be the most important points. In addition, an internet search of some points in the article has not been able to be verified. The resulting article would have been better if the topic had been broader and broader written, such as “Sustainable Supply Chains”.
Artificial intelligence is good at making predictions. Most forms of AI are predictive machines. ChatGBT is a statistical sentence/paragraph completion machine that composes content based on probabilities. The purpose of ChatGPT is not to tell the truth, but to write articles that look like they were written by a professional writer. The danger with generative AI is that the articles it generates can be a patchwork of lies.
ChatGPT can obviously be used to create training manuals. But Manhattan Associates’ Ciotta believes the potential of this technology goes far beyond that. Manhattan’s R&D team has been using ChatGPT for several months. They were using this technology long before it was in the news. Ciotta understands the potential role generative AI can play in improving Supply Chain applications. Warehouse management systems generate a lot of data, which is essential for AI to work properly. He imagines a manager asking his WMS, “Who are the top 3 pickers today?” Or “Who should be assigned to the arrivals dock?”
He also believes that ChatGBT can be used to configure WMS. If the bot has been trained to recognize all configuration options, a supply chain manager working in a retail chain could give the machine a new view that includes inventory from all fulfillment centers and all stores capable of fulfilling parcel shipments. You can ask to create to the customer! “Similarly, you can use a bot to test the generated code.
Polly Mitchell-Gutherie wrote an excellent article on this in Logistics Viewpoints. For generative AI to reach its potential, it needs to ask bots the right questions. There is an emerging field called Prompt Engineering that focuses on this task. Human judgment must be applied to the output.
A colleague and I spoke with Manhattan CEO Eddie Capel. Of course, we asked him when this kind of functionality would be included in future product releases. he didn’t know To use this technology, he must obtain a license from his parent company, OpenAI. They don’t yet know what the license will cost or what restrictions the license will have.
follow me twitter. check out my website.
