AI, supercomputers can tackle some of the world's toughest challenges: HPE's Justin Hotard

Conversations about the impact of artificial intelligence (AI) are everywhere – from shopping malls to factory floors and from conferences to boardrooms. This is especially true with the emergence of generative AI, a type of AI technology that can produce various types of content, including text, imagery, audio, and synthetic data. TechCircle caught up with Justin Hotard, who leads HPE’s High-Performance Computing (HPC) AI business group, at the HPE Discover 2023 event in Barcelona, to discuss issues like responsible AI, gen AI's business advantages and risks, advances in supercomputers, and more. 

Hotard expressed excitement about large language models and generative AI, stating that these advancements will bring new sources of productivity and value, particularly in scientific and technological breakthroughs. He gave an example of U.S-based Recursion Pharmaceuticals, an HPE customer who uses AI to analyze cellular microscopy images and find personalized medicine or identify new diseases. Hotard also mentioned a researcher in London who uses AI in neurology and brain scans. He finds these specialized use cases particularly exciting because of their positive impact on the world. However, Hotard believes that while AI offers many opportunities to enterprises, it also requires a complete rethink of how businesses operate. 

Hotard believes that it is the responsibility of technology providers to equip companies with the necessary tools for safe AI practices. He compared it to the emergence of firewalls in the early days of the internet. Hotard emphasized the need for tools that ensure companies can use AI responsibly. These tools include data security and validation, ensuring accuracy and comprehensiveness, addressing issues related to confidentiality and copyright, and removing bias from the data. 

Read more: We’re open to meaningful acquisitions that add value to our portfolio: Antonio Neri, HPE CEO

Hotard said that companies still have a long way to go in terms of deploying safe AI practices. He stated that companies should be able to audit and explain models, retrain them when necessary, and fine-tune them for better performance. HPE is investing across the ecosystem, building foundational models and developing solutions like Gen AI Studio to make it easy for enterprises to work responsibly and effectively, mitigating risks and delivering value to customers. Hotard noted that while data lineage, data versioning, deployment, and explainability are crucial aspects of safe AI practices, there are additional layers to consider as AI continues to evolve and become more integrated into business environments. 

Hotard also shared some best AI adoption practices for CIO/CISO in the coming months. He emphasized the importance of data integrity as the foundation for successful AI adoption. This involves ensuring the validity and robustness of the data, as well as proper labeling. Hotard also highlighted the significance of modeling and testing before deploying an AI solution. Thoroughly testing and evaluating effectiveness, potential bias, and desired outcomes is essential. Hotard stressed the need to monitor the deployment and usage of AI systems, gather feedback, and seek partners specializing in explainability and monitoring within the ecosystem.

Moreover, as the world becomes more complex, high-performance computers with AI give scientists, engineers and business technologists a powerful way to tackle some of humanity's most challenging problems. This is where supercomputers come into the picture. Supercomputing is a type of HPC that leverages a powerful computer to reduce the overall time to solution. HPE which recently invested over $22 million in a Global Center of Excellence for High-Performance Computer Manufacturing Center in the US, believes HPC and AI are a force multiplier and that companies can use HPC and AI to drive competitive advantage. 

Read more: HPE is refining GreenLake strategy with hybrid cloud push: CTO Fidelma Russo

Hotard believes that supercomputers are pushing the boundaries of solving the world's most challenging problems. For example, they are using supercomputers to make progress in tackling the challenging problem of nuclear fusion and optimizing green energy, such as wind farm optimization for maximum efficiency. These initiatives are meaningful and can have a significant impact.

Hotard sees supercomputers and AI as time machines that provide visibility into future technologies and predictive modeling scenarios. He finds it exciting to think about merging AI and energy to solve energy problems, like nuclear fusion. Fusion can help design more efficient airplanes and vehicles and power autonomous and robotic solutions. In the field of life sciences, supercomputers can assist in battling diseases and studying genetics with AI. Hotard believes these advancements are within reach in the next few years with supercomputers. 

On the debate about whether supercomputers are smarter than humans, Hotard said, today, computers currently run algorithms and mathematical models, with humans in control of them. But we should be thoughtful about the protections and limitations of humans with technology, as humans can be destructive with technology, as we have seen in the past.
“Nonetheless, we are still a long way off from computers being completely autonomous and humans not being in control,” he concluded. 

[The writer was in Barcelona to attend HPE Discover 2023 hosted by Hewlett-Packard Enterprise (HPE)]