Whitepaper: The Future of Skills in the Age of AI
By Thomas Crabtree, Academy Director at Corecom Tech Academy
Introduction
I’ve worked in and around Software Testing for just about my whole career. For those who remember the early hype around test automation and, although not directly related to testing, robotic process automation (RPA), the story probably sounds familiar. We were told that vast amounts of roles, especially manual testers, administrative workers, operations analysts, and a few others, would soon become obsolete. Selenium / WebDriver, which leads the charge in browser automation, and then the emergence of various other similar tools, was going to replace people doing manual testing with scripts and bots doing the work instead. RPA was going to do away with administrative workers and save the time of lots of other people.
The impact? People were going to be replaced by technology.
Has that happened? Certainly not for software testers, as it’s still one of the main roles in software development teams. The use of test automation in development teams is extremely high, but testers are still a fundamental role in most development teams.
What about RPA, which is still a relatively new solution? Adoption is high among larger businesses, with lots of roles, mature tool offerings (like UIPath and Power Automate), and solutions in place to streamline processes and reduce waste / costs, but back-office functions haven't been replaced with robots.
Did test automation and RPA completely change the skills landscape across the industry? Maybe to some degree, but not in the way it was predicted.
Instead, we’ve seen a more nuanced (and arguably more powerful) approach – automation optimising how we work, streamlining testing and back-office functions, and freeing people to focus on quality and problem-solving. Teams became more efficient, not smaller.
As artificial intelligence (AI) enters the mainstream in a similar wave of hype, it’s important we look back at those lessons. AI is, and will be, transformative, but like automation before it, it's real power lies in enabling people to do more valuable work, not necessarily eliminating the need for people.
In this article, we’ll examine what AI means for the skills of the future.
Lessons from Test Automation and RPA
The Promise of Automation
Test automation and RPA emerged to tackle repetitive, time-consuming tasks. For test teams, this meant automating regression tests, freeing testers from hours of manual clicking and checking. In finance and HR, RPA bots can handle repetitive data entry and document processing. The sales pitch: let the machines do the boring stuff, and humans can focus on higher-value work.
The Reality
But the outcome wasn’t widespread job elimination. Companies still hire testers, analysts, and various roles in back-office teams. What changed is how these professionals work. With automation handling the grunt work, testers now spend more time on exploratory testing, usability reviews, accessibility testing, and continuous quality improvement. RPA empowered operations and back-office staff to become process experts and improvement advocates, not just task executors.
In both test automation and RPA, automation didn’t eliminate roles, it shifted the demand toward higher-value knowledge work.
Automation didn’t replace humans, and it didn’t reduce the number of available jobs. There may be some disruption while changes happen, but not to the scale often claimed. It changed workflows and expectations, but not the fundamental need for skilled people. AI is following the same pattern, but with broader reach.
The Current AI Landscape
We’re now at a point where AI is being adopted by lots of businesses across the world, and across different industries. Unlike test automation and RPA, the impact is going to be much larger, and adoption is happening much quicker. With test automation and RPA, adoption has taken several years, and especially with test automation, the impact has been gradual, and the changing skills landscape is still happening now.
Knowledge Work Optimisation
The biggest leap has come from generative AI - large language models (LLMs) like ChatGPT, and tools like GitHub Copilot. These tools can write code, generate text, summarise documents, and even create media. In many cases, they deliver almost working solutions or proof of concepts almost instantly. The recent hype around ‘Vibe Coding’ is probably not a realistic solution for most large businesses, but the time saving from using Generative AI can be significant.
But they still need a human in the loop. We don’t replace the engineer or the author - we speed up their first drafts and help them think differently about how that progresses into a workable solution.
In terms of skills, users need to learn how to use Prompt Engineering effectively, review critically based on knowledge, and iterate. It’s rare that initial solutions provided by Generative AI are sufficient.
Decision Support AI
AI is increasingly being used to analyse large datasets and surface insights - recommendations in CRM systems, summarised customer sentiment, trend predictions. These tools help you make decisions faster and, with the right context and background knowledge, better decisions. The interesting thing with Decision Support AI is that much of the decision making that this supports just wasn’t happening before – for example, analysing customer trends in real time just wasn’t possible, or analysing large and fragmented datasets just wasn’t seen as a practical solution.
In terms of skills, the ability to interpret AI-generated insights in context becomes key, as well as the development and deployment of these new solutions.
Intelligent Agents and Task Automation
A newer emerging idea is the use of AI Agents, but much more sophisticated than chatbots. Although starting to sound like robots taking our jobs, these are systems that can take actions across applications and systems, not just respond to questions like Generative AI. These solutions aim to automate workflows ranging from typical email / meeting type processes (e.g. automatically booking meetings) to more complex ecosystems of workflows and tasks (e.g. smart automated employee onboarding). This is where the novel idea of “digital employees” comes in, although the actual application of that premise is yet to be seen (without the need for just as much time invested in supervision).
There are aspects of AI Agents being used now, like help desks and HR Support, and the idea of chatbots is common. Expanding the idea of a chatbot to include decision making and taking action is where this emerging solution is heading.
In terms of skills, the deployment of AI Agents will be holistic in nature: it will be about integrating and chaining tools, understanding business workflows, educating business teams, managing security and user errors. Roles will be similar to RPA – workflow designers, tool integrators, operations / support.
Skills That Will Matter Most in the AI Workplace
Test automation and RPA changed the skills needed to operate in the respective teams successfully. They didn’t fundamentally change the goals and objectives of these teams, or make those teams disappear altogether.
Rather than creating a new workforce to replace existing people or deploying robots that negate the need for humans at all, AI is prompting a broader reshaping of what it means to be “skilled”. But this reshaping of skills is more urgent as the adoption of AI solutions takes shape much quicker than other emerging technologies.
Test automation and RPA changed relatively unskilled work into knowledge work, requiring the use of tools, programming, analysis, critical thinking, and communication. These people used test automation and RPA; they didn’t build it from scratch.
Initially, lots of people with deep AI Engineering skills will be required, but over time that may diminish in favour of people that can use customisable solutions and tools alongside traditional software development and existing systems and workflows. AI will change some knowledge work into “meta-level work” - where workers oversee, guide, and collaborate with intelligent systems, focusing more on system design, reasoning and continuous improvement.
AI Engineering
Today, AI engineering skills are essential for building, tuning, and deploying the core technologies that power AI solutions. These include expertise in areas like machine learning, data pipelines, model evaluation, and infrastructure.
Right now, organsations adopting AI at scale rely on these specialists to:
Train or fine-tune models for specific domains or tasks.
Build and maintain the infrastructure required to serve models reliably and securely.
Create pipelines that feed clean, relevant data into AI systems.
Evaluate model performance, mitigate bias, and ensure outputs are trustworthy.
Integrate models into software products and larger systems.
However, as AI platforms mature and off-the-shelf models become more powerful and adaptable, the skills required will start to shift. More businesses will adopt AI through prebuilt tools, APIs, and platforms that lower the technical barrier, meaning the skills required will become more holistic and in line with existing skills and techniques required for building complex systems with 3rd party integrations.
Deep AI Engineering skills remain foundational, but over time the value will come from how these skills are applied in context; blending software, systems, and business understanding to bring AI to life where it matters most.
Systems Thinkin and Workflow Design
Understanding how to structure problems and workflows so AI can be helpful is an essential skill, not too dissimilar to the skills required to deploy RPA. It goes far beyond using AI to generate text or code. It involves a deep grasp of your end-to-end processes, identifying bottlenecks, inefficiencies, and decision points where AI can provide the greatest impact. This means knowing how to:
Map complex workflows and pinpoint where automation or AI-driven insights can improve the completion of tasks.
Break down high-level goals into manageable steps that AI systems can understand and support.
Integrate AI outputs into existing human decision-making without disruption.
Continuously evaluate and adapt workflows as AI capabilities and business needs evolve.
This kind of systems thinking ensures that AI tools become seamless and useful, not just useless features.
Human Judgment and Critical Thinking
Despite the illusion of mastery, AI models still make mistakes and miss important context. Prompt Engineering has already become a key skill for anyone who uses Generative AI now, but this will expand as the tools and use cases expand. Key skills involve:
Spotting when AI outputs seem questionable or inconsistent.
Understanding the limits of AI’s knowledge and reasoning.
Knowing when to trust AI recommendations and the sources / biases that might be at play
Asking the right questions and probing AI responses to clarify ambiguities.
Combining AI insights with domain expertise, ethical considerations, and situational awareness.
In other words, critical thinking and sound judgment remain essential. As AI takes on aspects of knowledge work, these higher-level thinking skills become the new standard.
Communication
Effective use of AI depends heavily on how people communicate with it. This includes:
Writing clear, precise, and context-rich prompts to get useful AI responses.
Explaining AI-generated outputs, recommendations, or risks to stakeholders who may not be familiar with AI.
Documenting decisions made with AI support to maintain transparency and accountability.
Translating AI capabilities and limitations into actionable business terms.
Managing expectations around what AI can and cannot do.
As AI is adopted into real world solutions, AI ambassadors and leaders will need to provide clear understanding and interpretation to other people - strong communication skills are more important than ever.
Technical Literacy and Software Development
You don’t need to be an AI Engineer to succeed, but basic technical literacy is becoming increasingly important for most roles. This means:
Understanding how AI models work at a high level and the basics of data quality and data analysis
Being able to read, interpret, and critically evaluate AI-generated outputs or insights.
Knowing how to integrate AI tools into everyday workflows or systems.
Familiarity with low-code/no-code AI platforms to experiment and prototype solutions; taking this further with programming and broader software development and operations skills to fully build, integrate and run solutions
Collaborating effectively with technical teams who build or maintain AI systems.
Existing roles need to adapt to new AI tools and technology, from back-office / business teams needing awareness through to IT teams augmenting existing technical skills to include the use of AI tools and future platforms and solution integration. This isn’t a significant change, but it is essential alongside deep AI Engineering expertise.
Ethical and Risk Awareness
AI introduces new risks and ethical challenges that teams need to consider, such as:
Algorithmic bias and unfair treatment of individuals or groups.
Misinformation and “hallucinated content” that can mislead decisions.
Decision-making that’s hard to audit or explain.
Data privacy concerns and regulatory compliance.
Potential misuse or unintended consequences of AI systems.
As AI adoption scales, having people who can spot risks, challenge assumptions, and establish checks against problems is essential to build trust and sustainability. Ethics and risk management become core business capabilities, not after thoughts.
What This Means for Businesses and Teams
Team Size and Impact
AI may have some short-term impacts, but it’s about reshaping teams not reducing them altogether.
At the time of writing, some businesses and areas within businesses are seeing a dramatic reduction in junior hiring: for example, graduate level Risk Analysts in financial services and insurance are increasingly being augmented or partly replaced by AI solutions, especially where there are repetitive (like test automation) or pattern based tasks (similar to some RPA solutions). Fewer, more skilled people are therefore required in these specific roles, but these are partly replaced by skilled AI Engineers. There are similarities in legal, marketing, and professional services to name a few.
Other roles, like Software Developers and Business Analysts, could increase in numbers due to the impact of AI. For example, as AI becomes more common across businesses, they’ll be a greater need to bridge the gap between business needs and technical solutions, especially where the business doesn’t have a greater depth of AI awareness, increasing the need for Business Analysts.
Re-skilling
Forward-looking businesses are investing in AI-ready talent pipelines now. Roles are evolving: with test automation for example, testers needed more technical, programming, and tool-related skills. With AI, testers will need more skills in areas like model validation, data quality analysis, critical thinking, prompt design, and the ability to evaluate AI outputs for accuracy, bias, and edge cases.
Across Software Development teams, being able to work alongside AI Engineering solutions and AI Engineers is a key part of successful adoption. To build Decision Support AI into an eCommerce platform means the eCommerce team needs to be able to design, build, test and deploy solutions that integrate AI Engineering professionals and solutions without causing delays and problems.
Talent Pipelines
Hiring AI Engineers for a specific solution or proof of concept also needs to take into account the broader skills teams now need, highlighted in this article, across all areas of the business. This may look like increasing the number of graduate or entry-level roles available (for example in Software Devlopment), to provide a pipeline of AI literate and skilled talent.
Other skills like adaptability, curiosity, communication, critical thinking and systems thinking need to be increasingly valued over narrow technical skills or people with a specific degree. While technical skills are undoubtably important, and technical literacy will be important across most roles, broader thinking and communication skills will become more important generally.
More than ever, a clear hiring strategy that includes all aspects of skills required for AI adoption is critical.
Conclusion
AI is the next stage of the same journey we’ve been on with automation for years. Like test automation and RPA, its real power lies not in removing people, but in elevating what they can do.
The workforce of the future won’t be smaller - it will be smarter, faster, and more skilled, precisely because AI takes the mundane and simple tasks away. Businesses that embrace this shift will help define it.
Get in touch to learn how Corecom Technology Academy can help find the perfect solution for your business.