{"id":35113,"date":"2024-02-11T10:51:25","date_gmt":"2024-02-11T10:51:25","guid":{"rendered":"https:\/\/digileaders.com\/?p=35113"},"modified":"2024-02-15T11:21:33","modified_gmt":"2024-02-15T11:21:33","slug":"the-secret-to-ai-the-universe-and-everything-learn-to-ask-better-questions","status":"publish","type":"post","link":"https:\/\/digileaders.com\/the-secret-to-ai-the-universe-and-everything-learn-to-ask-better-questions\/","title":{"rendered":"The Secret to AI, the Universe, and Everything: Learn to ask better questions"},"content":{"rendered":"
\n

I needed cheering up. With so many dark and disturbing stories in the news these days, I decided that I should escape for a while and remind myself of happier times by spending a few hours on the sofa under a duvet re-reading Douglas Adam\u2019s wonderful series of \u201cHitchhiker\u2019s Guide to the Galaxy<\/a>\u201d books. Taking me away from today\u2019s pressing problems and into another world where anything and everything is possible.<\/p>\n<\/div>\n

\n

Cult classics when they were originally written in the late 1970\u2019s, it was great to reconnect with the crazy characters and their adventures in Adam\u2019s books. Although to be honest, some of the material (or perhaps it is just me) has not aged too well. What seemed fresh and original more than 40 years ago is much less so today. Nevertheless, the central theme of these works and the genius of their key premise remains as fresh and captivating as it ever did. And in a new age of AI, it is perhaps even more relevant today than it was all those years ago.<\/p>\n

 <\/p>\n<\/div>\n

\n

Deep thought’s dilemma: Why the Answer is just the beginning<\/h3>\n<\/div>\n
\n

In Douglas Adams’s “The Hitchhiker’s Guide to the Galaxy,” the supercomputer Deep Thought famously calculates the answer to the ultimate question of life, the universe, and everything:\u00a042<\/a>. But, as the story unfolds, the real challenge turns out to be not the answer itself, but rather the question that Deep Thought spent millions of years processing. In fact, figuring out the right question to ask is far more critical than simply seeking the perfect answer.<\/p>\n<\/div>\n

\n

This sentiment resonates deeply with my experiences when considering the current state of AI. We’re bombarded with data, impossibly complex algorithms perform billions of calculations, and mountains of academic papers and reports promise new insights. Yet, there are worrying reports that\u00a0data management practices are out of control<\/a>,\u00a0finding meaning in the vast data sets is getting harder<\/a>, and\u00a0\u00a0many AI projects underperform<\/a>, failing to deliver the transformative results expected.<\/p>\n<\/div>\n

\n

Perhaps, like Deep Thought, we’re focusing too much on the answer and neglecting the crucial first step: asking the right question. In the world of AI, success hinges not on finding the “42” in our data, but on meticulously crafting the questions that unlock its relevance, meaning, and potential. It may be this critical shift, exploring how to focus on better questions, not just better answers, that can pave the way for more meaningful AI success in your organization.<\/p>\n

 <\/p>\n<\/div>\n

\n

Data-driven everything<\/h3>\n<\/div>\n
\n

With the rapid advances in AI, the allure of data-driven decision-making is undeniable. In an era where information flows freely, excitement surrounding AI focuses on the ability to quantify and analyze what previously seemed obscure and unmanageable. AI seems to hold the key to unlocking optimized solutions in almost every domain we can imagine. However, a singular reliance on \u201cgetting the data\u201d can create a mirage of clarity, masking the critical role of context in interpreting and applying information effectively.<\/p>\n<\/div>\n

\n

Hence, while much of the AI hype may indeed turn out to be justified, those of us involved in complex digital transformation scenarios also recognize that there are many potential pitfalls of these data-centric approaches if we don\u2019t acknowledge the fragility of decision-making when the bigger picture is neglected.<\/p>\n<\/div>\n

\n

Consider the example I faced recently when working with a multinational organization implementing a data-driven inventory management system across its global supply chain. The algorithm, optimized for efficiency, had been delivering great result in streamlining production and reducing stockpiles. However, as the context in which it operated became less consistent or predictable, it failed to account for local variations in demand, infrastructure limitations, and cultural nuances in distribution channels across the globe. The result? Delays, stock shortages, and ultimately, dissatisfied customers. This stark scenario brought home to me the crucial role of context in understanding data and its implications. They\u2019d been focused on the wrong questions.<\/p>\n<\/div>\n

\n

The rush to \u201cdata-driven everything\u201d must be aligned with an important reality: Data, in its raw form, is merely a collection of facts. It is the interpretation and application of these facts, informed by a nuanced understanding of the surrounding circumstances, that yields truly valuable insights.\u00a0As we have seen time and time again<\/a>, ignoring the context \u2013 the cultural factors, logistical realities, and unforeseen variables \u2013 can lead to seemingly efficient decisions that unravel in the face of real-world complexities. Nowhere is this being seen more clearly than with Generative AI tools based on Large Language Models (LLMS) trained somewhat indiscriminately on data pulled from across the internet. Real-world data, especially text and images scraped from the internet, is\u00a0riddled with bias<\/a>, from\u00a0gender stereotypes<\/a>\u00a0to\u00a0racial discrimination<\/a>.<\/p>\n<\/div>\n

\n

To address this, digital leaders and decision makers require a deeper understanding of the limitations of data-driven decision-making in isolation. They must delve into the root causes of data misinterpretations, and develop practical strategies for incorporating context into the decision-making process. In particular, they should realize that data is a powerful tool, but it is only when wielded with an understanding of the bigger picture that it leads towards informed and sustainable solutions.<\/p>\n

 <\/p>\n<\/div>\n

\n

Why start with why?<\/h3>\n<\/div>\n
\n

A good place to start in understanding how to use data is to recognize that the key question in any data-driven scenario is not to focus on \u201cwhat\u201d or \u201chow\u201d, but to start with \u201cwhy\u201d. Something that was highlighted some years ago in the work of business guru\u00a0Simon Sinek<\/a>.<\/p>\n<\/div>\n

\n

Although originally targeted at much broader business strategy concerns,\u00a0Simon Sinek’s call to “start with why”<\/a>\u00a0holds immense relevance in the data-driven landscape of AI. His argument centres on the idea that people connect with purpose, not just products or services. Businesses that communicate their “why” \u2013 their core values, beliefs, and motivations \u2013 cultivate deeper relationships with customers, employees, and stakeholders.<\/p>\n<\/div>\n

\n

In the context of our AI-driven digital age, “starting with why” translates to understanding the purpose behind data collection, management, and analysis. Going beyond mere data acquisition, it emphasizes extracting meaningful insights to solve real problems and create positive impact. Whether in\u00a0smart buildings optimizing energy use<\/a>,\u00a0connected cars<\/a>\u00a0enhancing road safety, or\u00a0wearables enabling personalized healthcare<\/a>, the critical challenges for digital strategy involve much more than numerical analysis and statistics. They require interpretation of that data in situations that are frequently\u00a0volatile, uncertain, complex, and ambiguous (VUCA)<\/a>.<\/p>\n<\/div>\n

\n

Consider any dataset that you are using today. I would argue that using that data to drive automated decision making in an AI scenario could be viewed as irresponsible without recognizing basic concerns such as;<\/p>\n<\/div>\n

\n