Artificial Intelligence (AI) is an umbrella term that covers many high-dimensional statistical methods such as Deep Learning or Differentiable Programming. These methods can be used in various ways to improve the operational performance of supply chains. However, both problems and solutions differ vastly from mainstream AI problems such as Natural Language Processing (NLP).
Artificial intelligence is a subject that has become somewhat of a buzzword in the world of technology over the last year or so. It has made numerous headlines, with certain experts claiming that AI will replace half of the jobs in the world by 2050.
In this episode of LokadTV, we try and rise above some of the buzz and instead focus on AI’s applications. We discuss what these applications are likely to look like in the real world and how they could revolutionise the supply chain industry.
However, we can safely say that many supply chains are still in the dark ages - too frequently managed by Excel sheets and run in a very manual, human driven way. We estimate that it’s going to take decades yet before they fully embrace Artificial Intelligence. AI, Machine Learning and Deep Learning would work very well for supply chains however, due to the mass of data at hand and the fact that processing power is so much cheaper, more performant and more readily available nowadays.
Most actual supply chain challenges lie in edge cases, which take a lot of manpower to solve, and this is a problem that AI would figure out more efficiently - for example, if you’re a fashion company, being able to forecast how many people are likely to return products. But we’re not all Google, Facebook or Amazon, so how can regular companies employ experts in AI? Well, the solution is actually to outsource to another company (shameless plug: a company like Lokad)…
To conclude, we discuss how AI does have its downsides, as it forces companies to become more “rational” and often challenges a business’ status-quo, completely changing budgets, roles, etc., which is why it is no easy thing to implement.
00:35 When do you think AI will come to the supply chain industry? What sort of time frame are we talking about here?
02:39 What can supply chain practitioners expect from some of these advanced statistical methods? Is there anything more than “better forecasts” that they can expect?
04:05 You are saying that there are problems that, on the surface, do not look at all like a statistical problem, which can in fact be treated like a “forecast” of some kind to make it possible to use those AI technologies. Perhaps you could expand on this a little further?
06:30 What can AI do for supply chains in general?
08:18 Having AI detect some of these edge cases seems like a fantastic idea. However, what will this actually look like? Will it be something similar to Siri or Cortana?
11:41 How much AI expertise is going to be needed to make a project like this work?
14:47 Will the stress caused by AI be a factor which blocks, or at least slows down, the adoption of the technology by companies?