Artificial Intelligence or AI has been cast as both a superhero with mystical powers and a villain with evil abilities capable of both taking over the world and people’s jobs. Is it the enemy, or even the end, of human intelligence or HI?

Let’s put it in perspective. The same polarised panic greeted the arrival of the motor car on roads in 1901.

Food packaging was one of the first areas of industry to embrace the powers of AI accessories, robots, in the 1980s. There was no polarised panic, although robots took over tasks previously used by humans, such as pick and place, case packing and machine slicing. Robotics was embraced as more efficient, more cost effective and more able to produce the volumes needed in a modern world.

AI is a far more powerful tool than these simple robots. And a fast learner. Presently, though, it still depends on a human programmer. The quality of an AI’s work depends on the quality of the request from a human. Human intelligence is essential for an AI program.

Human intelligence and AI are already collaborating effectively in FMCG. Most organisations do descriptive and some predictive analytics but not many have mastered prescriptive analytics.

Here the predictive AI models are used to either simulate many possible outcomes, or possibly run optimisation models across a complex set of possible decisions. The outputs of these processes are either fed back to a decision maker in the business to enhance their effectiveness. In a few cases AI makes the decision automatically on their behalf.

AI and human intelligence enhance one another

One of the next big things we see on the horizon is the use of these large language models (LLMs), taking things beyond the one-on-one conversation and acting as agents that effectively answer user queries and business challenges with decision support based on company data.

The challenge of extracting insights from data, is that data is typically stored in a rigid structure (tables, columns, rows) in order to maintain its integrity, but this rigid structure isn’t in line with how people think and communicate (pictures, stories, anecdotes). One of the oldest challenges in software engineering is, how do you take a user query and translate that into a meaningful request that a database can respond to that answers the question.

Large language model agents are potentially able to reason and autonomously execute tasks, acting as the go-between for the user and the company data, able to understand the intent behind the user’s request and simultaneously find the right raw data, transform it and perform the analysis required to meet that user request.

Doesn’t this application have the potential to supercharge productivity and decision-making effectiveness, unlocking the key insights that hide in the data we are storing? Doesn’t it assist human intelligence, that can then act on the information to create better business outcomes?

AI and human intelligence work well together

How could AI fit in suppliers’ future? Imagine a world where you’ve spent your summer account planning for the following year and you’re preparing to go and sell your plan to your retailers. Imagine if you had a negotiation expert built into your TPM, that you could feed the priorities for your buyer and tell that expert negotiator how they typically behave.

Imagine if that expert had full visibility of your plans. Imagine if you could ask it what it thought the likely objections might be. Imagine if you could ask it the best way to defend against those objections. Imagine this supplier’s future.

What about the future for retailers and partners? Both are always looking to cut costs and become more efficient. What if they had an AI twin, one who knows all the guidelines of what they can and can’t approve. Who knows the personality of the human it represents, and what their walkaway positions are. And because you can create as many AI twins as you like, they can negotiate with 2,500 suppliers simultaneously. Imagine this retailer and partner’s future.

Will AI solve all of our problems? Conversely, will it cause them? In the short term, there will certainly be problems as human intelligence grapples with capability, governance, processes and legal changes. Every tool we’ve ever created had to be tweaked. The motor car still is.

In the long term, AI should grow to become a tool that allows more effective decision making, conveying its learning to human intelligence, who will still make the go or no-go decision.

Human intelligence will continue to be as important as artificial intelligence as we need to motivate teams, inspire change and build capability with effective change and people management almost always being the blocker to change that conquers technical challenges. In the long term, the team of human intelligence and AI will realise feats that we humans can only imagine in 2024.

Max Lawday is head of Asia Pacific at Acumen.