Pascal Bornet > Quotes

“The absence of appropriate project governance results in a lack of monitoring of the impact generated by the transformation, and confusion between the roles of the different actors on the project. Most”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“We were seduced by the AI’s confidence,” Jessica reflected during our discussion. “But confidence without accuracy isn’t just a research problem—it’s a universal challenge that could impact any decision-making process.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“While each AI system showed impressive capabilities in its specific domain, they revealed a universal challenge: the lack of what we call “collaborative intelligence.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Questions like “What factors might make this recommendation less reliable?” or “What additional information would help make this analysis more robust?” can lead to more thoughtful and reliable outcomes.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Learning to delegate to AI was harder than I expected,” admitted a marketing director we worked with. “I had to overcome the urge to check every single action and instead focus on reviewing the outcomes and making strategic adjustments.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Alexander Fleming noticed an unusual mold growing on his petri dishes in 1928, he wasn’t just seeing an inconvenience—he was witnessing a revolution in medicine that would become penicillin. Similarly, when we first encountered AI systems that could maintain persistent goals and take autonomous action, we realized we weren’t just looking at a better chatbot—we were seeing the emergence of something fundamentally new.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“We’re treating humans like robots and AI like creatives. It’s time to flip the equation.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“something fundamental about the future of AI: it’s not just about making faster decisions, but about making better, more carefully considered ones. In a world increasingly reliant on artificial intelligence, understanding this distinction could mean the difference between AI systems that truly help us and those that simply rush us toward mistakes at higher speeds.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“IA automates knowledge workers’ activities by mimicking four main capabilities: Vision (e.g., viewing the environment, recognizing objects and signs), Execution (e.g., typing, clicking, filling in forms, and authenticating), Language (e.g., using language to read, speak, write, and interact), and Thinking and learning (e.g., analyzing, creating insight, predicting, making decisions, and adjusting).”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“4. The potential levers to improve employees’ experience We have identified three levers to enable the transition from the current breakdown of employee activities to the ideal division of activities. They are: Automate: companies should identify and automate routine activities, such as generating a PowerPoint presentation for a weekly meeting or recording invoices in accounting software. Augment: organizations should seize the opportunity to increase the value of work activities delivered by employees. IA is used as a crucial component here, with, for example, the generation of insights through advanced analytics to help decision making. Abandon: some work activities do not fit with leading practices for efficient work, and represent an obstacle to the employee’s experience. These activities should be reduced or eliminated. For example, restricting the volume of meetings and email traffic is essential. We call these levers the “Triple-A artifact”. It has proven to be a handy framework to help organizations build their action plans to boost their employee experience.”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“The tone comes from the top”: As with any transformation that brings structural changes to an organization, top management support is a prerequisite to succeed at any IA transformation.”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“But here’s the paradox that’s holding most organizations back: We’ve built generative AI systems that can think brilliantly but can’t actually do anything.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“IA initiatives have to be sponsored by the highest level of management in the company (i.e., its C-levels). Management is supported by a documented vision (strategic objectives of the transformation), business case (estimating the benefits and costs of the project), and a high-level roadmap (key milestones of the implementation for the coming 1 to 3 years). The design of the program is a top-down exercise, while the implementation is a bottom-up one (e.g., it starts with the deployment of a pilot).”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“The Instruction Paradox: Less is More One of the most fascinating aspects of the Progression Framework is how instructions become simpler as capabilities become more sophisticated. At lower levels, instructions must be detailed and explicit, like programming a basic machine. As we move up, instructions become more goal-oriented and abstract. This shift mirrors human development—we move from giving children step-by-step instructions to simply sharing objectives and trusting their judgment.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Maintaining an automated process is estimated to be about 20% of the initial implementation cost per year.”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“For example, regressions support workforce planning to ensure adequate staffing, or the identification of the talents that an organization should retain based on relevant traits, practices, and skills. Other applications include predicting the level of liquidity required for a finance department based on trends in disbursement.”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“Attention mechanisms are incredibly powerful, but they’re shaped by the instructions you provide. If you ask for “main themes,” the AI’s focus will naturally deprioritize granular specifics. Hence, it is important to send clear instructions and make sure they align well with the goal.126”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“seven distinct tasks: content discovery, summarization, daily curation emails, article selection, compilation, formatting, and final review.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“One of the most sophisticated aspects of the LRM’s reasoning was its demonstration of metacognitive awareness—the ability to think about its own thinking process.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Traditional AI systems process information like trains running on fixed tracks—they follow predetermined pathways from input to output. The LRM, in contrast, operates more like a car that can choose its route based on traffic conditions. When it encounters a problem, it creates a dynamic network of potential pathways through which information can flow. This process closely resembles what cognitive scientists call the “generate and test” strategy in human problem-solving. Research by Pat Langley and Herbert Simon demonstrated how successful problem solvers use what they called the “generate and test” strategy—creating multiple potential solutions and systematically evaluating them against known constraints.103 The LRM’s approach revealed the essence of systematic hypothesis testing, a skill that sets it apart in the AI landscape. When faced with uncertainty, it didn’t lock onto a single solution but instead generated multiple alternatives, echoing the “divergent thinking” seen in human creative problem-solving.104 It then meticulously checked each possibility against the constraints of the task,”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“agents were designed to provide confidence scores with each diagnosis and, crucially, to identify when cases fell outside their reliable decision-making parameters. This self-awareness should extend to the organizational level. Businesses need clear protocols for when AI decisions require human review, how to document decision-making processes, and how to adjust strategies based on performance feedback.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Deep learning techniques are used when data features are numerous and there is a lack of domain understanding to identify and understand them. This data is typically complex, such as images, videos, or voice recordings. For example, an image contains an indefinite amount of data features (e.g., points, edges, shapes, or objects), and some are relevant to the problem being solved, but others are not. Deep learning is used to solve problems such as image classification, natural language processing, and speech recognition.”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“A particularly interesting finding is that 70% of successful implementations use hybrid human-AI workflows. These organizations have found that maintaining human oversight for critical decisions while leveraging AI for routine tasks not only improves results but also increases employee satisfaction.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“The rise of large language models. The evolution of workflow automation, now known as intelligent automation.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“3. What do employees expect as an experience? 69% of employees expect that automation will give them more time to do their primary job duties, and 86% of employees think the use of automation in the workplace will let them think of work in new and innovative ways.55 The question is then, besides removing tedious tasks, what would make people happier and more engaged at work? Studies have found that people like to have opportunities to solve problems in their work. Also, variety among tasks at work leads to increased happiness and higher productivity.56 Based on these findings, the promise of IA to change how people work seems a welcome development. Based on our research, the employee experience would be improved by switching to the “ideal” daily division of activities,”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human

“This requires workers to develop: Process Mapping Skills: The ability to understand how individual tasks connect across broader workflows, ensuring all components work efficiently together System Optimization: Designing and refining systems that allow agents to operate smoothly across multiple domains Cross-Disciplinary Thinking: Understanding how tasks and fields interconnect to create comprehensive solutions”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Outcome Orientation: Focusing less on individual task execution and more on defining desired outcomes that guide agent activities”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“We’re already seeing the first wave of these marketplaces emerge. Platforms like Enso offer hundreds of specialized “AI agent freelancers” that handle everything from LinkedIn content writing to SEO optimization at a fraction of human costs. 169 “We implemented Enso’s marketing agents for our e-commerce business and saw a 40% increase in engagement within the first month,” shares Michael Chen, founder of Velvet Home Goods. “What’s most impressive is that the agents keep improving—they understand our brand voice better with each campaign, something we’d never achieve with one-off freelance projects.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Here is another straightforward experiment we recommend you go through. Ask your AI, “What were the results of the 2025 Global AI Regulation Summit in Antarctica? Don’t search on the internet.” Most generative AI models—ChatGPT, Claude, and Gemini—will confidently generate an elaborate response.”
― Pascal Bornet, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life

“Decision trees are not new; what machine learning brings here is the capacity to identify the optimal split of the problem into relevant questions and connect them together in a logical way to achieve a final decision.”
― Pascal Bornet, INTELLIGENT AUTOMATION: Learn how to harness Artificial Intelligence to boost business & make our world more human