AI and the ‘Connected Potato Syndrome’ or tips for a successful enterprise AI strategy

Executive Summary:

This document provides a strategic framework for integrating Artificial Intelligence (AI) into corporate environments, emphasizing the need to avoid the pitfalls of over-hyped innovations, as exemplified by the "Connected Potato Syndrome”? observed at CES 2020. It outlines a pragmatic approach to developing an AI strategy that aligns with business objectives and takes the best of AI responsibly and effectively.

Key aspects covered include:

Strategic Integration: Ensuring AI initiatives complement and enhance main business goals, with a focus on creating tangible value through improved efficiency, innovation, and customer engagement.

Practical Implementation: Guidance on selecting appropriate AI technologies adapted to specific industry needs, from healthcare to finance, and ensuring infrastructure readiness.

Ethical and Compliance Considerations: Addressing the importance of ethical AI use, including transparency, fairness, and adherence to regulatory requirements, ensuring AI solutions respect user privacy and organizational values.

Ongoing Adaptation: Encouraging an agile approach to AI deployment, allowing for rapid iteration based on feedback and changing market conditions to continuously refine AI strategies and solutions.

By adhering to these principles, organizations can support a productive, ethical, and forward-thinking AI culture that avoids the limitations of technology for technology's sake and instead focuses on strategic business optimization.

INTRODUCTION

Nicolas Baldeck tricked the Las Vegas CES 2020 selection committee by proposing a potato equipped with a microchip capable of "communicating" via Bluetooth with a smartphone application. At the time, I named this phenomenon, where the most extravagant concepts are presented as revolutionary innovations, the Connected Potato Syndrome.?

This syndrome perfectly illustrates the excessive expectations often placed on new technologies, stimulated by media hype, powerful marketing campaigns and, above all, the tendency we all have to exaggerate the capabilities of technology. All of this is taken up and amplified by social networks, driven by an unashamed ultracrepidarianism.

Ever since generative AI became the talk of the town, it would seem that the world is divided into three types of people when it comes to new technologies and, more specifically, Artificial Intelligence.?

1- The fans

First, there are those who adore anything remotely related to AI, even if it's just a connected potato.?

They're the first to express their admiration for the technology's "wow" factor. They applaud every announcement, no matter how magical, without worrying about petty details like "biases", "risks" or "limitations" or even the question "does it really work?". For them, every new product is a magic trick that deserves a standing ovation.

2- The Haters

Then there are those who hate everything to do with digital innovation. Opposed to decisions made by algorithms, they find everything useless or even dangerous, even before they understand the underlying principles or facts. Armed with a dose of skepticism, they listen attentively to speeches that limit and criticize. They suspect every technology of a conspiracy. For them, every innovation is a technological wolf in the human sheepfold, ready to devour society and take power over mankind.

3- The pragmatists

In the middle of these extremes, we find the pragmatists, those who seek to understand and learn. They don't let themselves be convinced by exaggerated rhetoric, but neither do they let themselves be disappointed by limiting points of view. They test, they evaluate, and above all, they seek to smartly integrate AI into their daily work lives without falling into technological paranoia or the euphoria of over-expectation.

Is Your AI-First Strategy Causing More Problems Than It's Solving?

“An AI-first approach could rapidly drive AI deployment across business operations, not because it solves “real” organizational or customer problems, but because AI implementation becomes an end in itself. The likely outcome is a lot of AI solutions in search of problems, or worse, solutions that create new problems” - Oguz A. Acar

In this Harvard Business Review article "Is Your AI-First Strategy Causing More Problems Than It's Solving?", Oguz A. Acar criticizes the current trend to prioritize AI in business strategies. He explores the implications of such a strategy, and above all stresses the need for a balanced, human-centered approach to integrating AI into business operations. He draws attention to the pitfalls that can arise:

- Prioritizing AI without a clear link to the organization's real needs can lead to the deployment of solutions that seek out problems or create new ones, instead of solving existing ones.

- Organizations can adopt advanced AI applications without having prepared the basic IT infrastructure, which can lead to failures in AI implementation and support.

- An AI-driven strategy can demotivate employees by suggesting that technology is more important than human staff, which can lead to lower engagement and resistance to AI initiatives.

- The replacement of human-driven roles by AI can lead to consumer negative reaction and undermine morale in the workplace, particularly when AI applications are perceived as detrimental to human interaction and teamwork.

- Focusing on AI deployment without considering ethical dilemmas and legal ambiguities can lead to decisions that prioritize technology over principles, with potentially significant negative consequences.

The article argues for a more thoughtful approach that prioritizes human and ethical considerations alongside technological advances to avoid the pitfalls presented by AI for AI's sake.

So how do you integrate AI into your business in the right way?

For companies looking to embrace artificial intelligence without succumbing to immoderate hype or total repulsion, here's a simple four-step approach. We'll then look at the views of experts such as Gartner, AWS, IDC and IBM.

1) Needs and Objectives Assessment

Before you even think about AI technologies, clearly define what you want to achieve in terms of your business strategy. From there, you can ask yourself what AI can solve, improve, optimize, simplify, reinvent (?) or simply automate. Here are a few examples from different fields:

- Healthcare: Personalize medical treatments, rapidly analyze medical images, manage health records, improve diagnosis and treatment.

- Manufacturing: Optimize production lines, improve predictive maintenance of equipment, automate processes to increase productivity and reduce costs.

- Finance: Increased fraud detection and automated portfolio management capabilities. Implement conversational chatbots to improve customer support.

- Automotive: Contributing to the development of autonomous vehicles, improving navigation systems, optimizing automotive supply chain management.

- Energy: Optimization of energy distribution, demand forecasting, integration of renewable energy sources.

- Transport & logistics: Logistics transformation with advanced fleet management systems, delivery forecasting and warehouse automation.

- Cybersecurity: Faster detection and response to security threats, by analyzing large volumes of traffic data to identify suspicious behavior.

These examples illustrate how AI can not only increase efficiency and reduce costs, but also uncover new potential for innovation and value creation in all areas of industry.

2) Choosing the right solutions

Not all AI tools and solutions are designed for the same purposes. On the contrary, they are specialized in specific fields. Some are perfect for predictive analysis, while others excel in task automation, or even image processing, text generation or complex calculations. For example :

- Predictive analysis: SAS Visual Analytics: Predictive analysis to forecast trends and behaviors.

- Task automation: UiPath: A Robotic Process Automation (RPA) platform that automates repetitive business processes.

- Image processing: Google Cloud Vision AI: Detects objects and faces in images for applications such as image classification and facial recognition.

- Text generation: OpenAI GPT-3: Advanced language model or Large Model Language (LLM) capable of generating coherent text for a wide range of applications, from content management to translation, for example.

- Complex calculations: TensorFlow: Machine learning framework developed by Google to perform the complex calculations required in deep learning algorithms.

- Voice recognition: IBM Watson Speech to Text: Speech-to-text conversion in applications such as voice assistants and automatic transcription.

Try them out, use demo versions, take part in vendor workshops, organize hackathons and, above all, choose solutions that align with the specific needs highlighted by your corporate strategy.

3) AI integration and deployment

- Iterative AI development: An agile approach is recommended for the development of AI solutions. The idea, for projects involving innovation, is to set up rapid iteration and improvement cycles based on user feedback.

This also enables us to check that artificial intelligence is being properly adopted internally. We need to measure the benefits and limitations of AI.

- Change management and training: Integrating and deploying AI in a company requires the introduction of new methodologies in line with a new vision. As with any disruptive project, this requires preparing the organization for change by clearly communicating the benefits of AI. This necessarily involves education. The success of artificial intelligence depends on team buy-in. It’s just as important to reinforce the effectiveness of AI technologies and methodologies by raising awareness and training teams as by adapting operational processes.

4) AI ethics and compliance

- Ethical considerations: Ensuring that the use of AI complies with ethical principles, including transparency, fairness, privacy and non-discrimination, is fundamental for a company.

- Compliance with regulations: Complying with laws and regulations applicable to the use of AI, such as the GDPR in Europe for data protection but also the IA Act helps to ensure the reliability of the project.

A corporate AI charter can facilitate the assimilation, adoption and monitoring of artificial intelligence within the company.

5) Measurement and adjustment

Once you've integrated artificial intelligence, you need to assess its impact in relation to your expectations and strategic objectives. You need to measure its benefits according to criteria such as efficiency, cost reduction and improved customer or employee experience, for example, using clear performance metrics.

Remain flexible to adapt your approach to technological developments. If the results don't meet your expectations, adjust your strategies, tools or objectives. An agile approach will facilitate rapid iteration and continuous optimization of your AI solutions.

WHAT DO THE EXPERTS SAY?

GARTNER

Gartner observes that many companies are actually in the process of exploiting use cases that adhere to their corporate strategy. Within this framework, their aim is to utilize the power of AI to improve productivity and strategic advantages.

Their document, "Map Your AI Use Cases by Opportunity", provides a framework for IT managers to define their AI ambitions, assess feasibility and prepare their organization to integrate AI solutions.

“‘Only 4% of IT leaders say their data is AI-ready” - Gartner

In particular, he stresses the importance of preparation in terms of cybersecurity, data management and he insists on the reasoned use of AI, thus ensuring that AI deployments align with broader business objectives.

Their study focuses on a number of areas.

- Mapping AI ambition: Gartner recommends that companies define and articulate their AI ambitions from day one. They suggest using tools such as the AI Opportunity Radar to visualize and prioritize opportunities.

- Three pillars of AI readiness: The success of AI initiatives relies heavily on implementing robust cybersecurity measures, ensuring AI-ready data and adhering to clear AI principles.

- Feasibility and integration: Feasibility assessments need to consider technical readiness, internal capabilities and external acceptance, and influence the choice between everyday AI applications and impactful AI innovations.

- Strategic alignment and continuous review: IT managers need to collaborate with their counterparts to align AI strategies with business objectives, constantly reviewing and adapting to the evolving AI landscape.

- Industry-specific use cases: The paper highlights the importance of tailoring AI applications to specific industry needs, improving operational efficiency and customer engagement.

Amazon Web Services

Let's continue our exploration of best practices for implementing AI into business strategy, considering the Amazon Web Services point of view. AWS provides a guide for companies as they develop their AI strategy. Beyond all the technical aspects, this document focuses on the business perspective of AI adoption. It highlights the potential of AI to help create new value, innovative products and services, and improve business results.

The document's development guidelines present :

- Innovate and accelerate: AI and ML capability can unlock new business value by enabling organizations to innovate at an accelerated pace and solve complex problems.

- Prioritization: The importance of having a clear AI strategy that focuses on tangible business outcomes and prioritizes high-value initiatives.

- Data Driven: The particular attention that needs to be paid to the effective management of data-driven and AI-infused products, model development and their management throughout their lifecycle.

- Collaboration and supervision: The adoption of AI within a company must be based on close collaboration and coordinated governance across the various branches of its organization.

- Innovation Management and Acceleration:? The impact of AI technologies on the market requires organizations to rapidly adapt their products and services to meet the expectations of customers, who have a very high level of expectation in view of the market push.

IBM

"Whether it’s deeper data analysis, optimization of business processes or improved customer experiences, having a well-defined purpose and plan will ensure that the adoption of AI aligns with the broader business goals" - IBM

On its side, IBM presents its perspective in a guide for companies looking to capitalize on artificial intelligence effectively. The article focuses on the transformative power of AI in various industries, detailing the benefits and steps needed to develop an AI strategy. According to IBM, a successful AI strategy must integrate with an organization's overall goals, offer a clear roadmap for implementation, and highlight specific areas such as data analytics, process optimization and customer experience improvement.

Key points include:

- Infrastructure et talents: Fundamental components of an AI strategy include developing the right technology infrastructure and acquiring talent. Organizations need to invest in the right solutions and skilled teams to effectively develop their AI initiatives.

- Ethical considerations: Ethical issues such as bias and transparency are crucial to the responsible deployment of AI. It's important to ensure that AI solutions are fair and compliant with regulations.

- Continuous adaptation: AI strategies must be dynamic, to enable companies to adapt to technological advances and emerging industry trends.

- Measurable benefits: The strategic implementation of AI can improve operational efficiency, foster innovation and enhance decision-making processes. In this sense, AI must have a significant impact on results.

IDC

"AI technology is capable of so many tasks. Organizations need to clearly define what element of their business they want to improve and tie their AI strategy to a business goal, and not a technology goal." - IDC

In its IDC white paper "The Four Elements Your AI Strategy Needs to Succeed" IDC presents a comprehensive framework for organizations aiming to effectively integrate artificial intelligence into their digital transformation initiatives. It emphasizes the importance of a structured approach, detailing four critical elements needed to build a successful AI strategy.?

- AI strategies should be closely aligned with the organization's overall business objectives, rather than purely technological ambitions, to ensure that they contribute to sustainable competitive advantages.

- The success of an AI strategy depends largely on the quality of the data used. IDC encourages organizations to base their strategies on rich, preferably proprietary data sets, supplemented by high-quality external data where necessary.

- Effective AI integration requires existing business processes to be updated to accommodate new AI technologies, hence the importance of change management and strategic realignment within the enterprise.

- Incorporating an ethical framework into AI strategies appears essential to combat bias and protect user data. It is vital to ensure that AI applications are responsible and aligned with organizational values and societal norms.

SUMMARY

Through these examples, we can see that experts from IBM, IDC, Gartner and AWS, agree that a good AI strategy requires alignment with business objectives, a clear understanding of the organization's infrastructure and talent needs, and consideration of ethical implications. They also stress the importance of data quality, adaptability and measurable benefits.

CONCLUSION

So, what’s the verdict? Well, whether you're watching AI evolve with interest, viewing every tech update with a dose of healthy skepticism, or navigating the middle path with rational thinking, the key takeaway remains: don't be blinded by tech’s bright lights without thinking it through. Sure, connecting a potato might grab headlines at CES, but is it what your business truly needs?

Think of AI not as a shiny toy to show off but as a toolbox filled with specialized solutions and methods. Each has the potential to fix something specific in your business or perhaps open doors to new possibilities. But it’s not just about having these solutions; it’s knowing when and how to use them that counts.

So, before rushing into AI implementation, take a step back. Assess your needs, align them with your strategic goals, and remember: the right AI application can drive your business forward, but only if you use it wisely. Keep your strategy grounded, your goals clear, and your approach ethical. That’s the way to ensure AI becomes a powerful ally in your business arsenal, not just a passing fad or an obvious mistake.

Let’s make sure that when we adopt AI, it’s not just for the sake of innovation but for meaningful, impactful change that propels us toward a smarter, more efficient future.

Sources & références : AWS: Business perspective: The AI strategy in the age of AI, Building AI responsibly at AWS?- Forbes: Somebody Snuck A Potato Into CES 2020, 15 Amazing Real-World Applications Of AI Everyone Should Know About - Gartner: Get AI Ready: Action plan for IT Leaders, Map Your Ai Use Cases by Opportunity?- Google Cloud Vision AI: Extract insights from images, documents, and videos - HBR: Is Your AI-First Strategy Causing More Problems Than It’s Solving?? - IBM: How to build a successful AI strategy - IBM Blog, AI Ethics | IBM, What is IBM Watson Speech to Text? - IDC: The four elements your ai strategy needs to succeed?- Llamasoft: Optimize your supply chain for an ever-evolving environment - OpenAI GPT-3:? OpenAI - SAS Visual Analytics: SAS Visual Analytics - TensorFlow: An end-to-end platform for machine learning - UiPath: Artificial intelligence for the real-world enterprise.