From Automation to AI Thinking Tools
AI thinking tools are systems that help people change how they frame problems, generate options, and test ideas, instead of only speeding up tasks that they already know how to do. That difference marks a new stage of cognitive automation, where the focus moves from repetitive chores to the structure of thought itself. In many workplaces, AI is still treated as an extra hand for drafting emails or summarising notes. But a growing group of workers is rebuilding their workflows around AI in more unconventional ways. They use models as searchable memory, sounding boards, and structured thinking partners. This approach does not replace their expertise; it stretches it, opening space for new questions and new forms of AI problem solving that would be hard to attempt within the limits of manual research and linear processes.
A Software Engineer’s Shortcut to Deeper Work
Software engineer and co-founder Williams Samuel used to avoid infrastructure projects because the research phase could take weeks of reading dense technical papers. Now he feeds those documents into tools like Google’s NotebookLM, interrogates them through questions, and filters down to what matters within hours. The task is not only faster; the shape of the work has changed. Instead of front-loading time into manual reading, he can spend more of his attention on design trade-offs and novel solutions. This kind of AI problem solving treats large models as scaffolding for thought, not a replacement for it. According to McKinsey, between 75% and 88% of organisations now use AI in at least one business function, but stories like Samuel’s show that the most valuable gains come when people restructure the work itself around AI-enabled exploration.
The Hidden Gap Between Speed and Transformation
Many organisations celebrate AI adoption once they see time saved on routine tasks, but efficiency alone can mask a deeper opportunity. Automation-first thinking treats AI as a digital assistant: compress meeting notes, draft standard reports, clean data. Transformative thinking treats AI as a platform for cognitive automation, where the goals include reframing problems, spotting patterns, and testing alternatives that were previously out of reach. The gap between those two modes is becoming a competitive fault line. Workers who only use AI to work faster keep their old mental models intact. Workers who use AI thinking tools to challenge assumptions rebuild how they plan, reason, and collaborate. This gap shows up not in hours saved but in the kinds of questions teams feel able to ask and the new options they can put on the table.
Cognitive Reframing as a Strategic Skill
Cognitive reframing with AI means asking models to restate, stress-test, or invert a problem so that fresh angles emerge. Instead of asking, “How do I speed up this report?”, high performers ask, “What different structure or metric would better describe this situation?” or “What assumptions am I ignoring?” These unconventional AI usage patterns turn models into sparring partners. By summarising divergent sources, suggesting counterexamples, or proposing alternative framings, AI tools help workers move beyond default solutions. The change is subtle but powerful: complex projects start with structured exploration instead of jumping straight to output. Over time, people learn to design prompts that mirror good critical thinking habits. In organisations that support this, AI problem solving becomes less about tools and more about building shared language for experimentation and discovery.
Building Organisations That Think With AI
To unlock the competitive advantage of AI thinking tools, leaders need to move past surface metrics like task volume and turnaround time. They can encourage workers to treat AI as a shared cognitive workspace: a place to map hypotheses, compare mental models, and record reasoning steps. That means rewarding experiments where teams re-architect workflows, not only automations that shave minutes off existing processes. Training should cover prompt design for exploration, not only templates for standard outputs. The goal is to make AI problem solving an ordinary part of strategy, design, and research, rather than a specialised add-on. As more workers rebuild their work around AI, the organisations that win will be those that recognise the gap between working faster and thinking differently—and invest in closing it in favour of deeper, more inventive thinking.