I was lucky to discuss ChatGPT with Polly Mitchell-Guthrie on Kinaxis’s Big Ideas in Supply Chain podcast. I’m a big fan of podcasts, so it was fun to be a guest. If you listen, I hope you enjoy it.
I learned a lot from Polly during the prep and the interview. Here are three lessons with a few additional links.
First, worst case, supply chain professionals should use ChatGPT as a “loss leader.”
Polly came up with this term. ChatGPT has made AI hotter than ever at the C-Level and with Boards. The graph below is from a recent Economist article showing the spike in “AI” mentions. I remember the spike we saw in ~2017, and I thought AI was hot then. It’s even hotter now.
Even if ChatGPT turns out to be a fad, we should use this moment to educate leaders about all the different aspects of AI and how this can help supply chains—using ChatGPT as the “loss leader” to start a larger discussion of AI.
Second, ChatGPT can help your supply chain team build more AI algorithms to unlock more value.
ChatGPT breaks down the syntax barriers of creating algorithms in Python (or other tools). This expands the number of team members who can prototype and test ideas.
We’ve been talking about citizen data scientists for a while, and P&G even talked about citizen developers within their supply chain.
ChatGPT might be the technology that unlocks this potential.
For more details, check out this Redmon and Davenport HBR article, 4 Ways to Democratize Data Science in Your Organization. They are frustrated by the lack of progress (this was written in 2021). But we may have the technology now to take a bigger step.
Third, can ChatGPT turn the vast supply chain data into something valuable?
Bloomberg released BloombergGPT to unlock the value of their financial data. Is there an equivalent for the supply chain?
I don’t know the answer to this question.
This leads to the last topic in this post: sharing about a research group I want to create.
I’m considering starting a research group to help understand and educate about how ChatGPT can unlock more value for supply chains (and businesses in general).
Some of the research topics would be the ones above— educating executive teams on the overall AI potential, enabling citizen data scientists, creating a factory for building supply chain AI algorithms, and understanding what SupplyChainGPT means.
The other big topic comes from a hypothesis that ChatGPT could help unlock (at scale) knowledge in academic papers, in specific exercises developed by professors, or in past student-industry projects. The idea would be to turn this into more accessible material for industry and education.
I’m still thinking about the details and talking to different people.
Mike, Agree with what you wrote. Buzz about ChatGPT has provided opportunities to data science professionals to educate supply chain leaders. Research group is a great idea for an emerging field such as Gen AI in supply chain. Let me know if I can help in any way.
Well said. I'd be happy to assist, support, or participate in a research group to support this exploration.
Also, you second point can't be emphasized enough. I have folks who are familiar with one language and not another language but are leveraging Chat GPT as a Rosetta Stone or code generator to build apps, transforms, or algorithms - syntax barriers are truly gone. Productivity and ingenuity is enabled by Chat GPT and other large language models.