How will the implementation of AI impact corporate location strategies?

The relentless evolution of artificial intelligence (AI) is likely to force companies to rethink their location strategies and site selection process, with revenue rivalling cost when deciding where to locate.

Introduction

The initial media frenzy about the impact of AI, focusing on jobs or swathes of entire sectors lost to automation, has subsided. Automation predates AI, and it will continue, but there is now recognition of the potential of AI to augment existing jobs and create new ones. Companies are now considering not just how AI can reduce costs through automation and productivity gains, but how it can support new ways of thinking that will challenge entrenched business models.

AI Clusters

Whilst technology skills are sought across all geographies and sectors, clusters have evolved where companies have aspired to:

• Attract the top talent, often irrespective of cost; or
• Identify locations where labour costs are low

The demand for top talent created tech clusters, such as San Francisco Bay and Seattle, where the largest technology companies are found thanks to the combination of the culture of innovation, talent supply from top universities, quality of life and venture capital.

Similarly, cities such as New York, London, and Paris and other major business centres also grew large technology clusters that extend beyond the tech sector into others such as financial services.

India remains the most scalable, low-cost destination for companies seeking tech skills, with Bangalore and Hyderabad rivalling or exceeding the scale of the US clusters. Whilst other low-cost countries seek to challenge India, it is difficult for them to match the scale of the talent pool, talent quality, graduate output, and the massive Indian IT outsourcing companies as a source of educated and trained talent. Labour costs are lower in some countries but the relative difference between them and India is insufficient to offset the other challenges of political and business environment stability.

Future Trends

Whilst India has grown to become the preeminent IT offshoring and outsourcing destination, there are challenges. The salary differential between high and low-cost countries is narrowing.

AI is not IT support. Outsourcing AI would risk the loss of competitive advantage. So, the question is, where will companies choose to locate?

Automation

The jobs most likely to be automated are those routine, repetitive, or basic analytical tasks that can be scripted and automated through AI and Machine Learning (ML). The process has been underway for some time. Contact centres have an increasing volume of non-voice interactions with customers, as people switch from phone to email, text, and chat bots.

Many of these jobs were part of the first wave of nearshoring and offshoring to low-cost countries and those countries are likely to see the highest headcount reductions.

Augmentation

AI will enhance human capabilities in tasks, such as data analysis, decision-making, and customer service. Again, there is already an ongoing evolution in these tasks, so perhaps the more exciting prospect is in areas that are not currently being considered.

Companies will see AI as a means of increasing revenue, rather than just reducing cost. AI will help people to think differently about their roles and how companies work. It will open potential new revenue streams through personalisation and may eventually lead companies to identify new products or services from the insights delivered by AI.

These jobs are in both high-cost countries within the HQ or close to customers and markets, and low-cost countries where work has been offshored. Whilst there will be productivity savings, it is equally important to consider how AI will also enable new ways of thinking.

New Jobs

The growth of AI has already seen the creation of new roles and growth of existing roles. Of those created, the highest demand has been for skills such as deep learning, machine learning, natural language processing, and algorithms. There has been significant growth in demand for data science skills and programming skills specific to AI, such as PyTorch. There is also an equal increase in demand for established programming languages such as JavaScript and Python to support the development of AI.

These jobs are being created globally, but there are proportionally more in high-cost locations where they sit alongside revenue-generating business units rather than being treated as a cost centre with a view to nearshoring or offshoring.

Clusters of AI Development Skills in Europe

The map shows the scale of the talent pool for 10 selected AI development skills, rather than AI users. These skills are concentrated in Western Europe, with London and Paris the largest clusters. Warsaw is the only city with an AI talent pool of significant scale in Eastern Europe.

University Origins of Recent Graduates with AI Skills

AI education is even more concentrated in Western Europe. Talent availability and the pipeline of graduates is two of the most important considerations for companies when choosing a prospective location. In these two alone, Western Europe outperforms Eastern Europe.

As mentioned earlier, this suggests that AI is sitting in HQ and customer markets, rather than low-cost nearshore locations.

The First ‘Reasoning’ Model

The recent launch by OpenAI of o1 represents a step change in AI maturity. The algorithm can have an inner monologue before it responds. Preliminary tests show that o1 is dramatically better at writing code, remaining coherent after writing hundreds of lines, unlike all previous AI tools.

Previous tools could adjust or extend existing code, but o1 can write entire modules or small applications. This has exceeded expectations and, according to OpenAI, the product has not yet reached its full potential.

So how will this impact on software development and its location? It may reduce demand for entry and mid-level engineers. There will be a shift in skills requirements, as engineers who can most efficiently formulate a problem for AI systems can reduce development time by days, even weeks.

Maximising the benefits of these systems will require a deep understanding of software development, creativity, and critical thinking, with a focus on these soft skills as coding is only part of software development.

This technology is still in its infancy, so it is difficult to predict the impact on location strategies. Labour costs in Europe are still significantly higher than those in India, but some reshoring is expected. This will depend on whether the productivity gains and benefits of proximity to decision makers and customers outweigh the increase in labour costs. If one European engineer can do the work of four in India, then it may be financially neutral. The higher the number displaced, the more likely the roles will be reshored to Europe and other high-cost countries.

Perhaps there will be a hybrid model, where using AI to build the core applications will take place in Europe, whilst more labour-intensive tasks such as customisation and localisation which require more human interaction and collaboration take place in low-cost countries.

The speed of this shift will be determined by companies’ adoption rates of o1 and the future versions of the model. Given the speed at which the latter is moving, companies need to start considering this now.

Future Trends

As companies continue to evaluate the potential of AI and the technology evolves rapidly, automation will lead to some job losses. However, it will also augment existing jobs, increasing individuals’ productivity, and the creation of new jobs.

Perhaps the most intriguing impact of AI on corporate strategy is the opportunity to process information in new ways that will challenge and ultimately disrupt existing business models and develop new products. The internet has had a significant impact on how we communicate and consume. AI has the potential to do the same.

The headcount required to develop AI tools may represent a small proportion of companies’ overall workforce. OpenAI employs around 2,000 people, of whom only 40% are technologists. Most of the employees are responsible for bringing the product to market.

Therefore, companies need to consider the other roles that can help them realise the benefits of AI. Critical in this are those who can interface between technologists and the business from AI integration engineers through to AI Ethicists.

AI will enable new ways of thinking. To deliver the best results, companies will need to identify locations that offer and attract creative, imaginative people who are not afraid to question often long-established business models. The challenge will be to identify the optimal location and spaces that can attract and retain this talent.

Perhaps the biggest question about AI relates to the ability to distinguish between fact and fiction, whether content is real or generated, or whether the internet is overrun by bots generating unverified content rather than humans. This has tended to be in the social and political domains, rather than corporate, where companies who are accountable to shareholders or regulated must base business decisions on reliable information.