Consultants need to understand how GenAI will impact new engineers

By Patrick Coleman

A hiker set off on a hike across the wilds. He relied on his portable global positioning system (GPS) device to map his trip. Part way his GPS failed. The story has two endings. In one scenario he pulls out a map and a compass and continues his trip, much more engaged in mapping his route. In the second scenario, like a handful of hikers last year, he didn’t make it home. The GPS is GenAI and the fate of the hiker is what this article is about.

WHAT IS GenAI?

John McCarthy, one of the founders of the discipline of artificial intelligence, coined the term “Artificial Intelligence” in 1955 to mean “the capability of computer systems or algorithms to imitate intelligent human behavior”. Artificial Intelligence became a branch of computer science dealing with the simulation of intelligent human behavior by computers, not a product.

Generative AI (GenAI) is artificial intelligence that can generate new content (such as images or text) in response to a submitted prompt by learning from a large reference database of examples.

GenAI is a form of Narrow (Weak) AI as it is designed to perform a narrow range of tasks with remarkable efficiency without operating beyond its programmed domain. The textbook ChatGPT Ate My Homework includes one chapter on what GenAI does well and one on what it does not. For example, GenAI is not good at original thought, eliminating bias, know-how in its limits, creating new empirical knowledge, applying empirical verification, taking courses (large language models are frozen), exercising morality, empathy or fear or eliminating hallucinations or biases. In fact, GenAI can be inaccurate and dishonest.

Before we panic, remember we are tool-using creatures. Using tools allows us to punch above our weight. We use tools, like GenAI, to amplify our expertise. We have time to sort out the issues because the adoption lags the hype.

WHAT WORRIES MENTORS?

We all work in teams with other humans. Teams facilitate the natural evolution from the apprentice to the master, creating a career path and sustaining a company reliant on knowledge workers.

Unlike previous iterations of technological breakthroughs, GenAI impacts our brain function, and the tool becomes part of the team. In a way, we become “cyborgs” or “MePTs”. The worry is, like the Star Trek Borg, can we survive being separated from the “collective” and think for ourselves? Do we know what cognitive skills are too precious to give up?

Mentors worry companies reward task performance over critical thinking. GenAI is unique as a tool since an engineer can perform without growing. Faster is not always synonymous with better (Ahmed 2025). AI accelerates, automates and optimizes, but what if that efficiency costs us depth, understanding and even identity? What if we start to produce “workslop,” AI-generated content that looks like work but lacks substance to move projects along?

THE RISK OF GenAI TO THE COMPANY

Mentors worry when a company focuses on rapid adoption of AI without developing a mature quality, risk and procedural ecosystem around the tool. A young engineer’s career can be ruined by a misstep with such a powerful tool. The World Economic Forum identified this as a major cyber risk in 2025, stating that “despite growing reliance on AI for cybersecurity, many organizations lack processes to properly assess the security of AI tools before deployment, creating a gap in managing associated risks.”

Insurance companies have similar concerns, warning coverage may be limited when GenAI is used. GenAI challenges fundamental concepts like authorship, ownership and understanding.

Engineers require a blend of traditional skills as well as knowledge of how to effectively manage and interact with AI systems. Firms with skill gaps are uncompetitive and vulnerable.

Companies must ask how GenAI adoption impacts succession planning. If AI makes an experienced person more effective, is AI then displacing the less experienced person who formerly did that role? Therefore, who will replace the experienced person when they retire or leave the firm?

Therefore, it is important to look before you leap, document everything well, train your staff and implement quality control procedures.

THE RISK OF GenAI TO THE EMPLOYEE

Brains atrophy if not used. For example, London taxi drivers, who have to memorize all possible routes across tens of thousands of roads to earn their license, have a larger and more developed hippocampus relative to London bus drivers who follow preset routes.

Research also uncovered a negative correlation between frequent AI tool usage and critical thinking abilities mediated by increased cognitive offloading. Younger participants are more vulnerable to this than older participants.

Researchers characterized what we lose when we outsource our thinking to GenAI:

• Critical thinking decline: Machines cannot reason through uncertainty. We need to think critically through ambiguity, evidence and contradiction. AI doesn’t just give us answers — it quietly takes away the need to ask better questions.
• Creativity erosion: Human creativity thrives on friction — on the uncomfortable space between clarity and chaos.
• Moral and intellectual laziness: AI does not weigh consequences. It does not have values, a conscience or accountability.
• Loss of agency: We stop trusting our own judgement as AI takes over the how, why and what of thinking. We fall victim to “automation bias.” When a tool speaks like an expert but reasons like a parrot, trust becomes a trap. Artificial intelligence is not artificial wisdom.
  GenAI, used incorrectly by those at an entry level in their career, rewards achievement over learning, diminishing the drive to think critically, be creative, and to stay current. Mental health can deteriorate because GenAI robs the user of a sense of accomplishment, undermines self-confidence and enhances the “imposter syndrome.” Tool dependence increases anxiety, guilt and lack of confidence.

Generalized cognitive atrophy is already a philosophical risk in itself. But there are other risks too. If this type of tool is widely used — and it is already the case with younger generations — we are at risk of a standardization of thought. British researchers showed that when authors asked ChatGPT to improve their work, the individual benefits could be great, but the overall creativity of the group reduced.

WHAT WE LEARNED ABOUT HIGH AND LOW PERFORMANCE GenAI USERS. WHO DID THE COMPARISON?

Three research teams studied high-performance and low-performance GenAI users. The consensus was that the high-performance groups use was balanced and comprehensive, primarily engaging with high-level cognitive processing, while the low performance groups use was unbalanced and fragmented, primarily engaging with lower-level cognitive processing.

Other takeaways from the studies are:

• Individual knowledge of the subject area increased staff performance and value-added actions.
• Without knowledge, staff became conduits for information, not creators themselves.
• While AI-assisted participants demonstrated enhanced task efficiency and reduced cognitive load, these benefits came at the cost of diminished critical thinking and a slight decline in self-efficacy.
• Integration of AI into cognitively demanding environments must be approached with caution, ensuring that users are not merely performing better but also thinking better.

The single behaviour that separated the groups is whether they used AI to amplify their expertise or to replace their judgement.

HOW DO WE STRUCTURE GenAI TRAINING BASED ON RECENT RESEARCH?

Lee et al state that “in light of these changes, training knowledge workers to think critically when working with GenAI should focus on developing skills in information verification, response integration, and task stewardship. Training programs should emphasise the importance of cross-referencing AI outputs, assessing the relevance and applicability of AI-generated content, and continuously refining, and guiding AI processes. Additionally, a focus on maintaining foundational skills in information gathering and problem-solving would help workers avoid becoming over reliant on AI.”

Yes, but this requires both actively engaging our critical thinking and continuing to exercise our neural pathways. AI can be a tremendous lever for intelligence and creativity, but only if we remain capable of thinking, writing and creating without it.

We need to always question the answers given by text generators and make a conscious effort to think carefully about what we read, hear or believe. We must also accept that reality is complex and cannot be understood with a few superficial pieces of knowledge. But the best advice is undoubtedly to get into the habit of comparing our point of view and knowledge with those of other people, preferably those who are knowledgeable.
Amlani and Davis propose in their book ChatGPT Ate My Homework a training structure based on a revised version of Bloom’s taxonomy (see Table 1 below).

The authors also identified four new competencies engineers must acquire:

• Melioration: Choose the right tool at the right time to complement our thinking and compensate for the tool’s shortcomings.
• Ethical reasoning: Decision about when it is appropriate to use AI and how we can appropriately use it.
• Collaboration: AI demands new skill sets, including prompt engineering, iterative conversation questioning, trusting outputs, and discerning our diverse viewpoints.
• Reflection: Demands that we consciously monitor and review processes as we work with GenAI.

EMPLOYERS AND RESPONSIBLE GenAI USE

Employers must ensure staff are not merely performing better but also thinking better. They need to reimagine entry points — collaborative role rather than a servant or a conduit role, encourage skill development for augmentation, not competition and provide a Human-AI workflow design.

We need mental friction in the workflow to build understanding and thought. Otherwise, we lose the ability to hold complex ideas in our heads, struggle through uncertainty, and earn our conclusions. For example, we no longer need to run to escape predators, but we still go to the gym because we need to exercise to remain healthy. We need to develop mental endurance by initiating the following:

• The most humane technology is one that slows us down. Introduce cognitive friction into your life.
• Set limits on AI use (e.g., use after a draft of a document or a plan is prepared).
• Reflect before prompting AI for an answer.
• Use AI-free time zones or days to reengage slow thinking.
• Cultivate deep work.

CO-PILOT, NOT AUTOPILOT

The KPMG Generative AI Adoption Index Survey assessed 2,606 Canadians (aged 18+) from August 6–21, 2024. 67% using GenAI do not think they are learning and retaining as much knowledge, 65% feel they are “cheating” and 82% claim GenAI-content as their own. These students will be our new hires.

Don’t just ask what AI can do for us. Ask what it is doing to our minds. We live in a world where thinking can feel optional. We can fall into the trap where we become emotionally detached and are unable to make decisions. “A mind that always has help nearby forgets how to stand on its own (Ahmed, S).” If this happens, humans stop being mentally vigilant and they become vulnerable to manipulation, misinformation and disruption.

Mentors stress that GenAI is best used as a co-pilot, not an autopilot. We need to do our own thinking and then use AI to sharpen it. Readers please note that GenAI was not used for this article.

Patrick Coleman, PhD, P.Eng., is a Principal Process Engineer, Wastewater with Stantec. Email: pat.coleman@stantec.com. Complete references are available upon request.

Bloom’s Taxonomy for GenAI.

1 — Identify
Engineers find and recognize relevant information, data and outputs. They know what tools give what output. They understand how to prompt to obtain the output they require.

2 — Interpret
Engineers can deduce meaning, limitations, and quality from AI output. They understand how GenAI works, so they can spot errors and shortcomings.

3 — Integrate
Engineers know how to combine AI input/output with traditional sources such as books, design guidelines, standards, and lived experience.

4 — Evaluate
Engineers have a sharp skeptical approach to assessing quality, credibility, relevance and verifiability of AI output. They know how and what to fact check, assess relevance, spot hallucinations and eliminate them, and understand the relationship between prompt quality and output veracity.

5 — Reflect
The engineer is competent at inward scrutiny and thinks metacognitively about their own learning biases and relationship with AI. They can recognize when they have become overdependent and need to backtrack.

6 — Collaborate
The engineer has strong teamwork skills, not only with AI tools but also with other humans. They contribute constructively and actively co-create solutions, make decisions and content. They develop an ability to negotiate, question, reframe and reach consensus, identifying common ground and working around disagreements.

7 — Judge
The engineer instinctively and genuinely executes ethical decisions about which tools to use, when and the proper uses of outputs. They have internalized values coping with authority, bias, learning outcomes and psychological safety.

8 — Curate
Engineers can draw out the best in themselves and AI-generated content. A curator produces work that is defensible, truthful, original, ethical, and properly referenced.

This article appeared in the Nov.-Dec. 2025 edition of Environmental Science & Engineering Magazine.