Subscribe

The Future of Innovation Is AI-Powered and Human-Led – THISDAYLIVE

7 minutes read
0 Views

Follow THISDAYLIVE on GoogleAdd us as a preferred source to see more of our stories in Google

Follow us on Google

By Shiro Theuri, CTO at Glovo

In record time, artificial intelligence has developed into one of the defining technologies of our time. In this still relatively early stage of adoption, much of the discussion and utility has been focused on productivity, automation, and questions around the extent to which AI will completely replace the need for human involvement. 

However, what is becoming apparent is that the real AI opportunity lies in utilising the technology to remove barriers that prevent humans from doing their best work. Used correctly, AI has the ability to amplify human creativity. 

Increasingly so, we are seeing AI make execution cheaper, freeing up human engineers to expand the scope and ambition of what they are able to take on. As a result, human input moves towards defining issues, setting direction, identifying trade-offs, and evaluating outcomes. 

For leaders in the technology fields, this requires a mindset shift. AI is a tool for writing simple code faster and implementing engineering tasks, but that’s not all it is. It should be viewed as a way of creating work conditions that allow engineers to work with more freedom, creativity, and innovation. 

Moving Beyond Productivity

Engineering is a far more creative endeavour than many people realise. Of course, product development requires code to be written, but real innovation happens as a result of problem-solving, experimentation, and endless micro-decisions, all of which are entirely dependent on human judgement.

Before the emergence of AI, developers often spent an unproductive amount of time on repetitive administrative tasks. Be it boilerplate code, sorting through documentation, debugging familiar issues, or navigating legacy systems.

AI already thrives by taking on many of these repetitive activities, allowing engineers to spend their time on work that creates genuine value. Think designing better architectures, improving CX, or developing entirely new product ideas. 

However, increasingly, AI is also able to take on complex engineering work, beyond repetitive, administrative duties. For instance, we’ve seen cases of AI suggesting root causes for engineering issues that helped my team avoid spending hours testing other potential solutions. As these capabilities ramp up, the onus for humans will shift to ensuring use of AI is safe and maximised, whilst carefully monitoring outcomes. 

The benefits are twofold. If an engineer is spending all day responding to tickets and doing repetitive development work, they simply don’t have the opportunity to explore new approaches or to experiment with possible superior solutions. Secondly, as AI takes on traditionally human-led engineering work, human governance and supervision become skills that allow engineers to thrive and develop. 

Hence, the new metric for success should be whether teams have the ability to innovate and manage AI, and not merely how quickly or how many lines of code are produced. 

The Role of Leaders 

Better technology by itself doesn’t create better engineering teams. The right leadership does. 

Onboarding teams to make the best use of AI is about instilling confidence and independent thinking. There is genuine and understandable fear among many engineers that AI may be here to negate the need for their skill set. The truth is, engineering roles are going to change, and the ability to complete certain tasks will become less valuable.

Given this, teams need to embrace a new reality by experimenting with AI to understand where it adds the most value, while prioritising skills which now have an elevated level of value, such as judgement, problem framing, system design and validation. The human-in-the-loop design model will become the new normal, and those that learn to oversee it effectively will be best placed to thrive in the years ahead. 

In this changed technological landscape, the organisations that succeed won’t necessarily do so because they have the best AI tools in operation, but rather because they’ve fostered environments that empower people to use them effectively and responsibly. 

Why Human Expertise Absolutely Still Matters

Regardless of the unquestionable brilliance of AI, it is still just a tool. It can generate code, suggest solutions, and take on certain engineering tasks. But there is plenty it can’t do that humans can. However brilliant the prompt may be, it cannot fully understand customer context, navigate shifting organisational priorities, or make nuanced product decisions that incorporate all technical and commercial considerations gathered from years of experience and human interactions in the field. 

Even in areas such as code production, AI is far from faultless. Data hallucinations occur, code can introduce security defects, and APIs that simply don’t exist can be invented. Only experienced, capable developers can spot these errors and address them before major damage is done. 

Software engineers still have a major role to play; it’s just evolving. They now need to be architects, reviewers, decision-makers, and problem solvers, using AI as a support mechanism for their work, without developing a dependency. 

For developers, it’s an evolution which needs to be embraced. In the direction we’re headed, developers will need to think more, not less. They’ll need to apply judgment, creativity, and experience to fix increasingly complex problems, while leaning on AI to handle execution. 

A Roadmap

Given some of the challenges described so far, leaders need to focus on empowering people as much as deploying technology. They can do so by first of all introducing AI to solve genuine challenges. By that, it means we need to improve how people work and the quality of their output, rather than just how quickly they work.

Secondly, there is a need to invest in capacity building by making the best, most effective use of AI. Training, experimentation, and knowledge sharing now need to become everyday occurrences.

Thirdly, accountability and judgement remain key for keeping standards high. AI-generated code should not mean a direct replacement. 

Lastly, the team must be given space to explore new ideas, improve technical foundations, and find contributions beyond their immediate remit. If AI succeeds in creating capacity and cheaper execution, the resulting time and resources need to be used wisely.

So much has been made of the potential for AI to replace jobs of all kinds, not only software developers. The resulting fear is understandable; we’re talking about people’s livelihoods. However, the fact remains that AI is a co-pilot, in need of a pilot. It can write thousands of lines per second and implement complex engineering work, but it can’t decide which problem to prioritise in order to address nuanced customer needs and human interactions. With no human guidance, AI merely builds on patterns that already exist. Real innovation can never result in this manner. 

The breakthroughs that will truly move the needle for businesses of all kinds will still occur as a result of humans asking better questions, making judgement calls, and, importantly, knowing when the machine has got it wrong. Further, human culture and discipline remain essential to shaping business practices, and can’t be dictated by AI. Only we as humans can navigate important business considerations such as workplace dynamics and ethical practices. 

The lasting focus as we learn to navigate this newfound two-way dynamic shouldn’t be on what AI can do independently from human supervision. Instead, it should be on what outcomes humans can create with the help of AI, whilst recognising when to trust it, when to challenge it, and when to ignore it altogether.