Although the use of artificial intelligence in business may not be new, the fairly recent popularity of generative AI platforms has thrust AI into the public consciousness. This, in turn, has led to companies hailing their utilisation of AI for a plethora of logistics and supply chain tasks.
Not all AI is the same, however, not only in terms of ability but also functionality.
This was very much emphasised during a recent discussion we had with Louisa Loran, Google’s Global Director of Strategic Industries for Transportation & Logistics.
With the AI hype train currently on overdrive, it’s high time many of us learned about the different forms of AI and how they can best be harnessed within a logistics and supply chain context. Therefore, Loran’s insights on the matter are as timely as they are valuable.
Generative AI
A logical starting point when it comes to artificial intelligence, is, of course, generative AI, which many consumers have already played around with. Millions of people around the world have now experimented with this tech to create a myriad of video, audio, imagery and texts. But we have only just seen the first small steps of its usage in enterprises.
So how can this technology be yielded to bring benefits in the logistics and supply chain space?
It is also going to play a role whereby chat interfaces are used to ask questions and get well formulated answers. Moreover, interfaces such as these may be utilised to ask questions of systems and databases. In fact, Loran notes that some of Google’s customers have already built such a data foundation.
What about the use of generative AI in a practical, everyday logistics concept, though?
Loran told Trans.INFO that the tech can be used to empower frontline workers in ways that were never previously possible.
“In the early days, we are likely to see this most in areas such as conversation and search. For instance, through that human-level interface, you can see someone at a factory or a port or whatever who may be able to, via their phone for example, gain insights on implications of decisions. Several layers of middle management will also be able to focus on more value-adding tasks than managing reporting. This allows businesses to empower front lines for faster decisions and brings greater alignment to any required changes.”
As Loran explains, historically, such a person either would not have access to that information, or they wouldn’t understand why that information was given to them, because it was more of a directive than an engagement.
At the same time, the person asking whether to roll a container or not can also feed into the human in the loop to understand what the crunches at the front line are.
In order to get the most out of AI, Loran also stresses that we are going to have to ensure we ask the right questions:
“If the person at the port asked about the fastest way to get home, the system would also provide an answer. That may or may not be the right answer for the business, but it’s a legit question to be posed from that person, and the system would give them an answer back. These models are being tuned, and you can tune them any kind of way you want, and you don’t want your employers to tune it to ‘How do I make my day as short as possible?’ I’d like to think that the desire to serve customers and build a profitable business should be higher, right?”
Understanding different forms of AI and avoiding generalisation
One unhelpful habit many people have nowadays, Loran notes, is generalising AI.
She emphasises that there are different kinds of AI, including more traditional AI like document AI, as well as vision AI, which helps to digitise the content.
In addition to those, there is also predictive AI for forecasting, as well as rule-based AI, which is used in operations research to optimise, allocate and schedule within confined constraints.
Finally, there is the aforementioned generative AI, which can complete sentences or translate across multiple modes.
As Loran explains, it’s not the case that you would deploy all of them in one way.
“One is more 1-1, one is generative of course, and another one allows a business leader for example to give some constraints for teams to then optimise within.”
The challenges and opportunities presented by the integration of AI-fluent future generations into the workforce
Another factor organisations will have to take into account when it comes to AI is how to drive competence across the generations.
While many of us are only just getting to grips with AI platforms, many young generation z talents entering the workforce are already well versed when it comes to generative AI in particular. The next generation, generation alpha, promises to be even more fluent in AI.
Such a high degree of AI competence will no doubt be a useful asset to companies. However, there are dangers and possible bad habits to consider here too.
So how can these young, emerging talents be harnessed to the benefit of an organisation as a whole, without them propelling the company in wrong directions as a result of not seeing the full context?
In Loran’s view, having an open mindset, both from top to bottom and bottom to top, will be absolutely critical.
In particular, it will be important for the young workforce of the future to challenge generative AI results in the same way we now cast a critical eye on media:
“I love the curiosity and critical nature of upcoming generations. However, I think there’s a risk that when they ask a LLM (Large language model) and it gives an answer, they may take that answer as the only truth. The reality is that their answer is down to the underlying model, the data set it was tuned on, how they posed the question, and so on. It’ll give them one point of view – just like the media often gives us one point of view,” said Loran.
Therefore, like many other areas, the value of being aware of one’s own perspective and being willing to see others’ perspectives remains essential, says Loran.
Leaders too will have to have to be on their toes and be inquisitive when required:
“At the same time, the senior leaders, who would likely know less about a specific issue, but have much broader context, will have to be curious to understand why this person is presenting a new input. So I think there will be a feeling of going slow to go faster, and time spent on reverse mentoring,” Loran told Trans.INFO.
The leaders who succeed, emphasises Loran, are the ones who shall maintain a sense of direction while being open to learning:
“Without the awareness of where we are venturing into, we’ll go around in circles. We won’t know where we’re going. At the same time, if you are not open to exploring, selecting, and importantly – deselecting, you are likely to be left at the station.”