Amazon's fixing the 'plumbing' of AI for businesses, says AWS AI chief Matt Wood

Summary

AWS Chief AI & Technology Officer Matt Wood explains new AI capabilities announced at a New York event, focusing on moving beyond efficiency gains to business reinvention by automating the 'plumbing' work between AI steps. He details how the cost of standard intelligence is falling rapidly, AWS's model-agnostic approach, the Nova model family, and the AWS Continuum cybersecurity automation tool. Wood also shares insights from his time at PWC on driving AI in large regulated organizations.

• AWS announces capabilities to move AI from efficiency to business reinvention by automating intermediate tasks.
• The cost of a given level of AI intelligence is falling by orders of magnitude every six months, while frontier intelligence remains expensive.
• AWS emphasizes model choice and agnosticism because different models excel at different problems.
• Amazon's Nova model family will remain competitive, with Wood expecting only 6–12 frontier model providers.
• AWS Continuum automates vulnerability validation, prioritization, and patching, reducing security team workloads.
• Wood's experience at PWC gave him empathy for AI adoption challenges inside large regulated organizations.