Developing an 'AI IQ' for Business Success

An opinion piece from the founder and CEO of Qantm AI and former IBM global chief AI officer

Seth Dobrin

October 2, 2023

5 Min Read
Abstract illustration of human profiles and a light bulb
Getty Images

Artificial intelligence (AI) holds tremendous promise. Yet, most companies have to work hard to capture its benefits. While data science and machine learning were hyped in the 2010s, fewer than 25% of firms adopted more than one AI model. Now, generative AI like ChatGPT is the hot new trend. For some reason, they believe they will succeed with generative AI using the same approaches. However, companies need to change their approach to avoid repeating past disappointments.

The solution lies in a proven methodology aligning AI to business strategy across people, processes, and technology using a human-focused and value-based approach. Here I outline a practical roadmap so organizations can leverage AI to achieve strategic goals from the C-suite to frontline operations.

Level-setting on the foundation

First, let us level set on the relationship between related technologies and the resulting limitations:

  • Machine learning (ML) is a subset of statistics enabling computers to improve at tasks through experience.

  • Deep learning is a subset of ML utilizing neural networks.

  • Pre-trained models are a type of deep-learning.

  • Foundation models are a type of pre-trained model.

  • Generative AI is a type of foundation model.

  • Large language models

Most people in the business world have taken a statistics class and understand that statistics is probabilistic. Since all of these technologies are based on statistics, they are all probabilistic in nature – 90% probably means that 10% of the time, the event won't happen. And probabilities compound. Two independent 90% probabilities used to predict one event together result in an 81% probability - almost twice as likely to give a non-ideal outcome.

As described above, ChatGPT-like models are built on deep learning foundations. These models are pre-trained on massive datasets -– think vast amounts of the internet –  to generate new data without needing lots of proprietary data. Applications include text, images, video, and more. But these models are probabilistic in nature, meaning as you chain more and more events together, the probability it is correct decreases. Today's generative AI models have addressed this to a certain extent, but the fact is that this probabilistic nature still impacts their efficacy, which causes some of the inaccuracies when combined with other inherent side effects of the underlying technology, so-called hallucinations.

Connecting AI to business goals

But for most companies, past ML initiatives delivered little to no value. Without addressing core issues like strategy alignment, capability building, and modernizing technology, generative AI risks becoming the subsequent overhyped letdown. The key is evolving the approach to AI.

AI strategy must directly support business objectives, not just deploy technology. A "strategy-first" approach ties investments to ROI and value creation to business impact. Data, AI, and generative AI are merely tools to accelerate reaching your business goals. The real question is not 'How can I use AI or generative AI in my business?' but 'What combination of technologies will help me reach these goals?' AI and generative AI will undoubtedly be part of the solution, but is your organization even prepared to use these tools, let alone implement them?

I refer to the execution of a human-focused, value-based approach as having 'AI IQ.' The reduction of this practice means having the knowledge, strategy, and execution capabilities for AI. A structured approach entails the following:

  • The board sets the vision and champions the value

  • Senior management operationalizing the tactics

  • Business units identifying where AI assists in decisions

Close collaboration ensures initiatives align with goals and smooth implementation. AI projects are prioritized by potential value and executed in agile sprints. This evolves AI from technology-first to business strategy-first.

Assessing AI readiness

Beyond technology, organizational culture, structure and processes determine AI success. AI readiness assessment evaluates the following:

  • Culture - promoting innovation, ethical AI, and continuous learning

  • Structure - dedicated AI teams, clear roles, robust governance

  • Strategy - aligning initiatives to goals, project management, scaling

  • Infrastructure - computing, storage, networking, software

  • Data - quality, governance, access, integration, culture

  • Skills - technical, domain, business acumen

Assessment identifies strengths, weaknesses, and areas to improve. It provides a roadmap to become AI-ready across the organization.

A 5-step formula to unlock value

A proven formula for maximizing value from AI investments:

  1. Create an AI strategy aligned with business goals

  2. Update governance and accountability structures

  3. Prioritize use cases based on value and impact

  4. Provide AI education at all levels

  5. Implement AI on use case-specific technology stacks

First, assess AI readiness to identify gaps. The strategy ensures alignment with objectives. Updated governance enables oversight of risks. Use cases are prioritized by value creation potential. Organization-wide education builds capabilities. Technology stacks are tailored to each use case's unique requirements.

This structured approach, akin to the formula used by successful studios like Pixar, unlocks AI's full potential across diverse applications.

Practical steps to get started

Here are three concrete actions organizations can take to start their AI journey:

  1. Conduct an AI readiness assessment - Identify strengths, weaknesses, and gaps across strategy, people, process, data and technology.

  2. Develop an AI strategy - Outline the vision, business goals and ethical principles. Define organizational roles and responsibilities.

  3. Start small, scale intelligently - Prioritize one to two high-value use cases. Implement agile pilots, learn and expand iteratively. Avoid boiling the ocean.

Key takeaways

  • Most companies have struggled to create value from AI due to issues like poor strategy alignment, lack of organizational readiness and ineffective implementation.

  • Generative AI opens new possibilities but will under-deliver without addressing these gaps.

  • A proven methodology aligns AI tightly to business strategy across people, process and technology.

  • AI readiness assessment identifies strengths, weaknesses and areas for improvement across critical dimensions.

  • A 5-step formula creates strategy, updates governance, prioritizes use cases, develops capabilities and implements tailored solutions.

  • With the right holistic approach, AI can help companies boost competitiveness by enhancing capabilities, serving customers better and achieving strategic goals.

Next steps

  • Conduct an AI readiness assessment

  • Create an AI strategy aligned to business objectives

  • Start small, scale intelligently

  • Develop organizational capabilities

  • Implement agile pilots tied to value

  • Continuously adapt as AI evolves

The time for action is now. Organizations can harness AI's immense potential to drive real competitive advantage with a focused approach.

Read more about:

ChatGPT / Generative AI

About the Author(s)

Seth Dobrin

Seth Dobrin is the founder and CEO of Qantm AI and the former chief AI officer at IBM.

Keep up with the ever-evolving AI landscape
Unlock exclusive AI content by subscribing to our newsletter!!

You May Also Like