Economist Impact - AI Compute Summit
UBSC at The Economist Impact AI Compute Summit
On Thursday, May 22nd, we had the pleasure of participating in The Economist Impact’s inaugural AI Compute Summit.
Christiane Ulriksen contributed in multiple roles — as an expert panelist in the "Keeping it Cool" session, and as a moderator for both a panel discussion and two fireside chats.
It was an inspiring day of dialogue around the future of AI infrastructure, sustainability, and collaboration across sectors.
Here Christiane participated as a panelist:
Keeping cool – innovations in data centre cooling
As AI compute grows to meet increasing global demands, cooling infrastructure is a critical challenge for data centres. Advanced cooling technologies are necessary for managing heat from high-performance AI workloads and are key to improving energy efficiency and reducing costs. From liquid cooling to airflow optimisation and beyond, which innovative solutions are most promising for the future of data centre design and operation?
Discussion topics:
What innovative cooling technologies are emerging to handle the heat generated by large-scale AI compute?
How can data centres balance energy efficiency, cost control, and performance through smarter cooling strategies?
What role do sustainability goals and regulations play in driving advancements in cooling infrastructure?
Following the panel, Christiane had the pleasure of moderating a fireside chat.
With Aline Almeida, Head of Research at IO.net Foundation, on the topic of decentralising AI compute.
The conversation focused on the emerging concept of GPU-as-a-Service — often described as an “Airbnb for GPUs” — which is rapidly gaining attention as a more flexible, distributed approach to meeting the growing demands of AI workloads.
The session welcomed audience questions early on, leading to a dynamic and engaging discussion.
Christiane also moderated a thought-provoking panel on:
Cloud resilience and the strategic choices companies face when considering multi-cloud architectures.
The panelists brought diverse perspectives to the table. A key takeaway:
For large enterprises like Nestlé or Kraft Heinz, a multi-cloud strategy is often essential for resilience, flexibility, and compliance across global markets.
In contrast, smaller, more agile companies such as Recursion may benefit more from a streamlined single-cloud approach, allowing for simplicity and speed.
Regulatory requirements, particularly for financial institutions and companies operating across multiple jurisdictions, further support the case for multi-cloud. As a relatable example for European users, GDPR was highlighted as a key driver.
The session underscored how size, sector, and regulation play critical roles in shaping an effective cloud strategy.
To close the day, Christiane had the pleasure of moderating a fireside chat with Donald Thompson, a Distinguished Engineer at LinkedIn/Microsoft, who joined live from Los Angeles via Zoom.
The discussion explored the transformative potential of Agentic AI — autonomous AI systems capable of performing complex tasks — and how they will reshape the workforce.
Key takeaways from the session included:
Agentic AI will fundamentally shift how we work, requiring HR teams to manage not just people, but also AI agents.
In the near future, professionals like IT developers may supervise AI agents that generate code, shifting their roles from creators to overseers and collaborators.
Donald also addressed the challenge of compartmentalization within large organizations. He described them as evolving organisms, often hampered by siloed thinking that obscures the broader impact of their decisions.
Closing Fireside Chat with Christiane and Donald Thompson: Agentic AI and the Future of Work
Unlocking business value: how AI compute powers competitive advantage