How you can exploit LLMs to drive human-like cognition into your business processes

Understanding how your business can exploit human-like cognition with Generative AI is a new business fundamental. If you aren’t clear on the outcomes you have set out to achieve, there is little chance you will be able to exploit it.

To uncover the true disruption of GenAI, look to its breakthrough technology – Large Language Models (LLMs). The secret to their success – and the engine that will drive the next significant enterprise disruption – is that they augment critical business capability – human-like cognition. Cognition is the mental action or process of acquiring knowledge and understanding through thought, experience, and the senses.

What an LLM is and how it delivers human-like cognition

LLMs are built with AI’s deep learning techniques to understand, generate, and manipulate human language. They are trained on and learn from vast amounts of text data from diverse sources.

This understanding, generation, and manipulation of language manifests as human-like cognition. In a human-like way, LLMs learn from data, make decisions, and use natural language to interact with humans and to support the automation of complex tasks.

While an LLM’s cognition is simulated and limited to pattern recognition and data processing (lacking the true understanding of actual human cognition), their simulation of human-like capabilities makes them well-adapted to handling the challenges of knowledge work, which often includes poorly defined tasks and unstructured data. Human-like cognition counts when handling ambiguous tasks – such as accurately interpreting what is meant evn f thr ar sm ltters mssg.

Today’s LLM-augmented CEO can make faster insight-informed decisions

We can use LLMs’ human-like cognition to augment our ability to make insight-informed decisions. They can handle many of the knowledge tasks or ‘moments of cognition’ once the preserve of white-collar workers.

For example, when deciding on a new product, a traditional CEO may turn to their insights people to have them apply their cognition to understand economic indicators, competitor activity, social media trends, and internal sales data. Teams will analyze that data to forecast market demand and customer preferences in a range of scenarios to come up with reports and recommendations for the boss. To deliver this takes hundreds of expensive ‘moments of cognition’ in which a human has had to make a decision based on their experience and information available to them.

LLMs are increasingly capable of handling many of these moments of cognition by continuously aggregating data from a vast array of sources – external and internal – analyzing for opportunity and risk, and applying predictive models to generate insight and recommendations. And they can do it all in seconds.

Conversations with your data

Today’s LLM-augmented CEO can query their data in natural language through an LLM: “Based on current market trends, what is the potential success of launching a new eco-friendly product line in the next six months?”

LLM: “Analyzing current market data and trends, the demand for eco-friendly products has increased 15% over the last quarter, particularly in the 25-35 age group. Competitors are also launching similar products, but there is significant opportunity in the mid-tier price segment, which is currently underserved.

LLM: “Recommended actions: Launch in Q3, focus on online marketing campaigns targeting environmentally conscious consumers, and highlight sustainable packaging as a key feature.”

CEO: “What risks should we consider?”

LLM: “There is a potential for supply chain disruption for sustainable materials and aggressive pricing strategies from competitors. Mitigate by securing multiple suppliers for sustainable materials and consider a tiered pricing strategy to remain competitive.”

Many of the moments of cognition required to deliver these answers to the boss can now be handled by an LLM. And more will follow. The ability to augment our cognition with the human-like cognition of LLMs is a new business fundamental with as much potential to disrupt as was caused by the web.

The web meant you didn’t need brick-and-mortar to trade. LLMs may mean you don’t need people in the loop

The web created value with connectivity (Exhibit 1). New business models were enabled by the connectivity it provided. The platform business model began to dominate markets. Platforms (in a business context) connect buyers and sellers. Previously, retail outlets and sales teams were your only viable route to market. The web changed all that. You no longer needed brick-and-mortar if you owned the digital platform to connect buyers and sellers.

Exhibit 1: Understanding the true disruption delivered through LLMs is critical if we are to be the disruptors rather than the disrupted.

Source: HFS Research 2024

In the metaphor of the ‘One-person Unicorn’ (exhibit 1), we reference OpenAI CEO Sam Altman’s assertion that AI will enable entire companies to grow managed by just one person – the goal owner. This individual would set rules and guardrails for AI that would otherwise operate with agency towards defined goals. There is widespread expectation in Silicon Valley that such a $1 billion company is not only possible (one such model is described in this paper) but that it will become a reality in just a few short years. Where the web meant you did not need brick and mortar to trade, LLMs mean one person can complete significantly more knowledge work. At the theoretical extreme, you could end up competing with a company that has next-to-zero people overhead.

We do not think this should cause panic, but it should guide your immediate roadmap.

As discussed in the HFS blog post, ‘We’re entering the Third Phase of AI: Purposeful AI’:

Imagine asking your digital assistant to onboard an employee, which automatically executes tasks such as device procurement, enrolling their benefits, setting up training, etc. Or demand planning, forecasting, and then optimizing a supply chain network at the click of a button? Or is your marketing AI companion setting up a podcast session for you, making suggestions on who to invite and how to promote over your social networks, etc.? Or a Soc-2 security audit that automatically delivers the governance plans and requests actions from your employees? These examples are likely not far off…

The One-Person Unicorn should be the metaphor guiding your roadmap

We don’t expect enterprises to become one-person bands overnight. But we do expect individuals to readily apply LLMs to scale what they can each achieve – combining their own cognitive capabilities with those of LLMs to deliver more effective moments of cognition throughout enterprise processes. Consider your whole team being augmented with LLMs and the multiplying power the technology will have on your outcomes.

The Bottom Line: Augmenting the power of cognition means every knowledge worker can create more value now

Where is the Amazon of GenAI? We cannot yet describe the new business model outcomes of the new GenAI paradigm – but we have identified that augmenting the value of cognition enables one person to achieve so much more than they previously could. Those businesses that grasp that disruption and put it to work fastest and with the greatest effect will be tomorrow’s leaders.

Everyone is a unicorn now.