Could AI help cure cancer? Why life sciences' next breakthrough could depend on data centers

Artificial intelligence could help find a cure for cancer within our lifetime, according to Rene Haas, CEO of global semiconductor company Arm.

But there is another part of Haas's prediction that could prove just as significant for the life sciences industry: having enough computing power to make it possible.

Speaking to the BBC, Haas said AI could eventually model biological processes too complex for today's technology, including how DNA markers are affected by cancer. But he also warned that AI's growth is already being constrained by the supply of chips needed for data centers.

For life sciences, the two issues are increasingly connected. As pharmaceutical and biotechnology companies use AI to accelerate drug discovery and analyse complex scientific data, demand for the computing infrastructure and specialist talent behind it is growing too.
The race to use AI in life sciences could therefore create a talent challenge extending far beyond the laboratory.

AI-powered drug discovery needs somewhere to run

Much of the conversation around AI in life sciences focuses on models, data and scientific applications. Yet increasingly sophisticated AI workloads also require significant computing infrastructure.

As demand grows, life sciences companies are broadly taking two approaches: investing in dedicated AI compute infrastructure or accessing capacity through cloud and technology partners.

Bristol Myers Squibb recently announced plans to expand its AI computing infrastructure with NVIDIA, giving its scientific and computational teams greater capacity to support AI workloads across oncology, haematology, cardiovascular disease, immunology and neuroscience.

Eli Lilly has similarly developed its own AI factory for pharmaceutical discovery and development. Other organisations are accessing the compute they need through cloud and technology providers rather than developing dedicated infrastructure themselves.

Whichever model they choose, AI-led research ultimately depends on the data centers, power, and infrastructure capable of supporting increasingly demanding workloads.

The talent challenge now stretches from the laboratory to the data center

That dependency has workforce implications too.

Life sciences organisations still require deep scientific expertise across oncology, genomics, molecular biology and clinical research. Increasingly, those teams must work alongside specialists in computational biology, machine learning and data engineering, supported by the infrastructure professionals responsible for delivering the compute on which their work depends.

Dave Bowers, Managing Director at EPM Scientific, says this shift is already appearing in hiring requirements:

Industry data shows 43% of pharmaceutical companies cannot find candidates with the digital skills their workforce now requires, while AI-related roles are regularly taking six to nine months to fill. On this, Dave says: 

At the same time, the talent market life sciences companies are hiring from is becoming much broader.

An organisation scaling AI-led drug discovery may require the following professionals:

• Computational biologists and machine learning scientists to develop models
• Data specialists to manage the information feeding them
• Cloud and HPC expertise to deploy workloads

And further down the infrastructure chain, data center specialists across areas such as electrical and mechanical engineering, cooling, power, commissioning and operations.

The future of AI-enabled life sciences could therefore depend on specialist talent across both scientific research and the physical infrastructure supporting it.

Pharma is entering a much bigger competition for technical talent

Traditional pharmaceutical businesses are no longer recruiting specialist AI talent solely from a pharmaceutical and biotechnology labour market. According to Dave: 

Compensation can make that competition particularly difficult.

Recent analysis found Anthropic offering approximately $300,000–$320,000 in base salary for a research scientist on its life sciences team, while a comparable senior machine learning position at Pfizer reaches approximately $166,500. According to Dave, Genentech is among the pharmaceutical employers that have moved closer to the compensation available within leading AI companies:

But competition extends beyond AI talent. Life sciences' growing reliance on AI also connects the sector to another talent-constrained market: data centers.

AI workloads depend on the power, cooling and critical infrastructure that specialist data center professionals deliver. As demand for compute grows, talent shortages in this workforce could create constraints for AI-led life sciences too.

Could data center access influence where life sciences companies build AI teams? 

Life sciences hubs have traditionally formed around access to research institutions, investment and specialist talent. As AI-led research scales, access to computing infrastructure could increasingly influence where teams are based too.

Dave believes this could become an important workforce consideration:

There are already signs of science, technology and infrastructure becoming more closely connected.

Novo Nordisk and AWS, for example, recently announced an AI co-innovation hub in London, bringing together AWS engineers and AI specialists with Novo Nordisk R&D teams, alongside a wider partnership making AWS the pharmaceutical company's preferred cloud provider and strategic AI partner.

For employers, location decisions could therefore increasingly involve assessing several interconnected ecosystems: 

• Where is the scientific talent?
• Where is the AI expertise? 
• And where is the computing infrastructure capable of supporting them?

From scientific breakthroughs to physical infrastructure 

Whether AI will ultimately help cure cancer remains impossible to predict. What is clearer is the scale of the ecosystem required to realise its potential, spanning scientific and computational talent through to the data centers and infrastructure powering AI.

For life sciences organisations, AI strategy is therefore becoming a workforce, infrastructure and location decision too. Dave summarises:

As investment in AI-led drug discovery accelerates, organisations need to plan for all three parts of the equation: science, compute and talent.

Planning the workforce behind your AI ambitions? Request a call back to discuss your hiring needs with EPM Scientific.