iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Before reaching for an AI tool, ask: Is this an AI-shaped problem? That question captures the central argument of George Lawton’s article: using AI well depends not only on technical fluency, but also on judgment, critical thinking, curiosity, collaboration and the ability to frame a problem. These capabilities matter when deciding what to ask, whether an answer is trustworthy and when AI is the wrong tool. The argument is a perspective illustrated by interviews and examples—not proof that human skills are objectively harder to learn than technical AI skills.
Why human skills matter when working with AI
AI can produce an answer quickly, but people still have to decide what is worth asking, whether the answer holds up and what to do with it. Caoimhe Carlos, Udemy’s vice president of global customer success, put the challenge this way: “How do we make sure that they understand not to trust implicitly, to ask the right questions of the technology? How do we make sure that the skills of judgment and discernment are present in our students and in our workforce of tomorrow?”
This is not an argument against technical training. It is a reminder that knowing how to use a tool is different from knowing when to trust it, how to check its output or how to connect it to a real need. Those decisions draw on human capabilities that are useful well beyond AI.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Start by defining the problem, not choosing the tool
Lila Ibrahim, described in Lawton’s article as Google DeepMind’s founding COO and its first chief AI readiness officer, emphasizes the value of asking better questions: “Getting answers is super cheap these days, right? But actually, knowing how to ask really good questions is getting harder and harder.”
#1 Best Overall
Her hackathon example turns that idea into a practical habit. Ibrahim said, “We spent the majority of our time not with AI but with a pen and paper, asking ourselves questions like, what are we actually trying to achieve? Is this an AI-shaped problem?” The team’s decision to spend time clarifying the goal before using AI is an anecdote, not a controlled test showing that this approach always produces better results. It does, however, suggest a useful sequence for anyone considering AI:
- Name the goal. Describe the outcome you need in terms that do not assume a particular tool.
- Identify what the task requires. Consider the information, judgment and human interaction involved.
- Decide whether AI fits. Use it when it can help with the task, rather than because it is available.
- Check the result against the goal. A fluent answer is not, by itself, evidence that the answer is useful or correct.
Check the foundations behind an AI answer
Problem framing also involves understanding the data a system can draw on. Oxford professor Michael Wooldridge challenges organizations to connect AI ambitions to the state of their information: “You want an AI strategy? Tell me what your data strategy is. Tell me where is the data that you want to build this AI? What form is it in? Are we using consistent notation? Is it all robust?”
Rank #2
His questions point to a practical distinction: an AI plan cannot be separated from decisions about where relevant data comes from, how it is organized and whether it is reliable enough for the intended task. For an individual user, the parallel is to ask what an answer is based on and whether that basis is appropriate before acting on it.
What Coursera’s 2026 figures do—and do not—show
Coursera and Udemy’s 2026 Global Skills Report covers learning data from more than 300 million learners across 98 countries and introduces an AI-Human Skills Synergy Index. Coursera says the index combines learning signals with national AI-readiness data. These are platform and index measures, not a representative census of each country’s workforce or proof of workplace outcomes.
Coursera reports 107% year-over-year growth in critical-thinking enrollments in its 2026 data. That figure describes enrollment behavior; it does not demonstrate that learners’ critical-thinking skills improved by 107% or that employers became more productive. The same report’s framing reflects growing interest in pairing AI capability with human skills, but enrollment interest and demonstrated ability are different measures.
Lawton’s article reports that the UK ranked 19th of 98 countries on the index. That is a ranking on Coursera’s AI-Human Skills Synergy Index, not a verdict that the UK’s entire workforce has mastered—or lacks—particular skills. The article also reports 126% growth in complex-problem-solving enrollments and says 75% of organizations were experimenting with or implementing AI while 18% said most of their workforce had been trained. Those latter figures are attributed to Lawton’s article and were not independently confirmed in Coursera’s report materials.
Coursera’s September 28 announcement also reports year-over-year micro-credential enrollment growth of 33% overall and 122% for AI topics. These figures describe enrollments, not course completion, assessed proficiency or transfer to work. The announcement separately confirms the 107% critical-thinking enrollment-growth figure. Read Coursera’s announcement of the 2026 report.
Collaboration helps bridge disciplines
Some AI work depends on people who understand different parts of a problem being able to communicate across disciplinary boundaries. Lawton points to DeepMind breakthroughs involving neuroscientists, ethicists and biologists as examples of work shaped by collaboration, not just technical ability. The lesson is not that every team needs the same mix of specialists; it is that AI-related problems can require perspectives and knowledge that no single technical role supplies.
Best Value
That kind of collaboration also depends on explaining assumptions clearly, listening to people with different expertise and being willing to revise a proposed solution. Those are human practices that help teams decide whether a system addresses the actual problem and what risks or needs might otherwise be missed.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Build skill through practice, not just exposure
Reading about critical thinking or communication is not the same as practicing it. When comparing ways to learn, consider what the learner actually does, who gives feedback and what the assessment shows:
- Practice: Does the learner work through realistic decisions, or mainly consume explanations?
- Feedback: Does it come from a person, peers or an AI simulation—and what can that source reliably assess?
- Assessment: Does it record course completion, demonstrate a skill or show that the skill transfers to work?
- Integration: Does the exercise develop human capability alongside AI use, or treat tool operation as the whole task?
Lawton’s article describes experiments with AI role-play for communication practice. A simulation might give someone a chance to rehearse a conversation, but its use is not evidence that AI role-play reliably improves communication or replaces the feedback and relationships involved in human interaction. It is best understood here as an experiment, not a proven solution. A critical-thinking or problem-solving workbook can offer optional practice prompts; it cannot substitute for useful feedback, collaboration or experience applying the skill.
Education initiatives are experiments, not verdicts
The article also describes UK lifelong-learning policy, Experience AI, a teacher-led program for pupils aged 11–14, and differing approaches to AI in schools. These examples show that AI education involves choices about who teaches, what learners practice and how institutions set expectations. The article provides no comparative evaluation establishing that one policy or school approach is superior.
Marni Baker Stein, Coursera’s chief content officer, describes why learning needs to keep adapting: “Jobs are still called the same thing. But what’s changing underneath them is the tools and the tasks and the workflows.” The tools used in a role may change even when its title does not, making it important to learn both how to work with AI and how to exercise judgment about its place in the work.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
