AI in Context: Operationalising Augmented Intelligence through the Data Lifecycle

The Turing Trap: Augmentation vs. Replacement

Erik Brynjolfsson’s “The Turing Trap: The Promise & Peril of Human-Like Artificial Intelligence” (2022) highlights a critical challenge in AI development: the pursuit of creating AI that is indistinguishable from human intelligence can lead to economic and societal risks if misaligned with human needs. Brynjolfsson explains that AI research focuses on automation by replicating human-like intelligence in machines, as measured by the Turing Test, rather than augmenting human capabilities. He describes this prioritisation of automation over augmentation as the “Turing Trap”, which can result in job displacement, economic inequality, and over-reliance on automation (Brynjolfsson, 2022). 

The key insight from Brynjolfsson’s paper for Defence and Security leaders is that AI should complement, not replace, human judgment. Instead of striving for autonomous systems that mimic human intelligence, investment should prioritise augmented intelligence—AI that enhances human decision-making, situational awareness, and strategic analysis. Humans bring unique qualities to data science by integrating qualitative approaches into augmented intelligence systems. 

As now explained, the data lifecycle provides a valuable framework to understand the adoption of augmented intelligence and the corresponding organisational transformation. In the context of the data lifecycle, the terms 'Digitisation', 'Digitalisation', 'Data Science', 'Artificial Intelligence (AI)', and 'Augmented Intelligence' are often used interchangeably. While these are related concepts, each plays a distinct role in understanding the adoption of AI. Misunderstanding these terms can lead to misaligned expectations, ineffective implementations, and resistance from personnel. Using precise terminology helps cut through marketing hype and focus on creating real operational value. 

The Data Lifecycle: A Foundation for AI 

Data lifecycles provide a structured approach to consider how data is collected, processed, and utilised to generate insights for informed decision-making. Various frameworks exist; the UK Government’s version, shown in Figure 1, offers a clear view of how to gain insights from data. A well-known operational example is the Intelligence Cycle, where data is collected, analysed, and transformed into intelligence. 

Using this framework, we can break down the following key concepts related to adopting AI. 

Digitalisation: Digitising Analytics Processes

Digitalisation applies to the entire data lifecycle and refers to using information technology to enhance organisational processes. Digitalisation is not just about converting analogue data into digital form; it involves the organisational transformation that results from adopting information technology. For Defence and Security, automating workflows, such as the intelligence cycle or processing seven key questions, or enhancing situational awareness using digital dashboards. Importantly, digitalisation focuses on the underlying methodology for how the end users analyse data to gain their required insights. 

Digitisation: Converting Information into Digital Form

Digitisation, a more specific term, applies mainly to the “collect, acquire, and ingest” and “prepare, store, and maintain” phases of the data lifecycle. It relates to converting analogue or non-computerised information into digital data. Digitisation is critical because high-quality digitised data is the foundation of effective AI systems. Poor-quality data inputs lead to unreliable AI outputs, so digitisation is critical to successful AI adoption. 

For example, in Defence applications, digitisation may involve converting paper-based reports into searchable electronic documents, translating voice communications into text, or using sensors to transform battlefield conditions into real-time data streams. 

Understanding the limitations of digitisation is also essential to understanding its limitations for human-centred AI. The quantitative and qualitative debate tells us that the underlying attitudes, beliefs and emotions of people are not really quantifiable in data. So effective human-machine teaming is required for human-centred analysis to uncover these attitudes, beliefs and emotions. 

Data Science: Extracting Insights from Data

Applicable to the “use and process” phase, data science encompasses the methodologies and methods analysts use to generate insights from data. In Defence and Security, data science applies to intelligence analysis through disciplines like Signals Intelligence, Geospatial Intelligence and narrative intelligence.

With the increasing adoption of AI, data science is too often characterised as a field for mathematicians or physicists who can code. The somewhat reductive characterisation of data science does not account for the qualitative aspects of data analysis, which the qualitative vs. quantitative methods debate has shown to be critically important. Moreover, data science does not have to include information technology; while not necessarily desirable, analysts can conduct data science tasks manually with pen and paper. 

Artificial Intelligence: Scaling Data Science with Algorithms

Also applicable to the “use and process” phase of the data lifecycle, AI is the scaling of data science over vast data sets. AI scales data science methods through software algorithms and large databases, thereby enabling the automation of complex tasks at a speed and scale beyond human capability.

For example, while a large team of human analysts might take months to analyse thousands of intelligence reports manually, an AI system can process and perform the same task in seconds. This ability to scale makes AI a powerful force multiplier for intelligence and security operations, shortening the time between observing a situation and making a decision. Notwithstanding, scaling data science should not ignore the integration of critical qualitative insights.

Augmented Intelligence: A Human-Machine Partnership

The concept of augmented intelligence is a more nuanced and practical approach to AI. While AI is often associated with replacing human roles through automation, augmented intelligence focuses on empowering analysts and decision-makers with algorithmically generated insights. 

Instead of replacing people, augmented intelligence provides advanced tools that enable humans to perform their jobs more effectively. This approach increases productivity and efficiency, allowing personnel to focus on tasks that require human expertise, which in the context of this paper refers to the more qualitative aspects of analysing data. 

For example, an AI-augmented intelligence analyst could utilise machine learning to sift through vast amounts of open-source intelligence while still applying human judgment to assess credibility and the underlying attitudes and beliefs of individuals. Similarly, AI can automate threat detection in cybersecurity, but human operators are still needed to interpret context, intent and make critical decisions. 

Conclusion

Brynjolfsson’s paper reinforces the need for qualitative methodologies in AI-driven analysis in which humans can add their unique qualities that AI can not replicate. By avoiding the "Turing Trap," Defence organisations can develop AI systems that empower human operators, leading to more resilient and adaptive decision-making frameworks. 

AI’s role in Defence and Security is rapidly evolving. Still, its success depends on how well organisations manage and integrate data across a chosen data lifecycle, such as the intelligence lifecycle. Digitisation and digitalisation set the foundation, data science extracts meaningful insights, and AI enables the extraction of insights at a large scale. Nonetheless, true operational advantage lies in augmented intelligence, where AI empowers rather than replaces human expertise. 

For Defence and Security leaders, the key takeaway is that AI is not a magic bullet but a powerful tool when integrated thoughtfully into human-centric workflows. Organisations can unlock AI’s full potential while maintaining trust and operational effectiveness by prioritising high-quality data, robust analytical methods, and human-machine collaboration. 

By embracing this human-machine partnership, Defence and Security organisations can enhance situational awareness, decision-making, and strategic agility—ensuring they remain ahead in an increasingly complex digital battlespace. 

References

• Brynjolfsson, E. (2022) The Turing Trap: The Promise & Peril of Human-Like Artificial Intelligence. Brookings Institution.

• UK Government (2020) The data lifecycle