Kirk Bresniker, HPE | HPE Discover 2024
summary
HPE Labs lays out five AI ethical principles (purpose, privacy/security, inclusive, human-focused, robust/responsible) and admits creating them takes a year not six weeks, then builds governance that asks for evidence of fitness: data lineage, energy used, and explainability. The team demonstrates an exascale data center digital twin for energy-transparent decisions, promotes direct liquid cooling backed by hundreds of patents, and pilots AI signals (transactional, satellite) to spot forced labor while warning that training a single large model can consume vast amounts of power.
AI-written summary · about
