What matters most for competitive advantage in AI infrastructure?

RF

Rob Ferguson

VP Technology and Strategy · Fireworks Ai

Data

Rob Ferguson says data matters most for building a competitive advantage, because the AI companies getting the most funding are data-focused and the best companies will really understand their data.

Sarah Sachs

Eng Lead, AI · Notion

Own Your Context

Sarah Sachs says companies should own their context; she argues the model should be evaluated on its own value, so they aren't locked into poor context or afraid to use a better model.

The answer starts at 8:43.

Amin Vahdat

Engineering Fellow and Vice President · Google

Specialize Hardware

Amin Vahdat says to specialize hardware to specific workloads, because specialization makes hardware faster and more power-efficient, though the workload must be durable enough to justify it.

Milind Naphade

Senior Vice President, Technology, AI Foundations · Capital One

Data Advantage

Milind Naphade says Capital One's AI advantage is its data advantage, describing the bank's approach as customizing the entire AI stack around its own data, policies and operating environment.

Vivek Vaidya

Co-Founder/CTO · Kana

Execution

Vivek Vaidya says the moat is execution rather than technology, asking whether companies are maniacally focused on execution.

Peter Desantis

SVP, AWS Utility Computing Products · Amazon Web Services (AWS)

Systems Approach

Peter DeSantis says AI infrastructure isn't a chip problem or a model problem but a systems problem, arguing it should be treated as such.

Peter Ludwig

CTO, Co-founder · Applied Intuition

Deploy Intelligence Fast

Peter Ludwig says companies will be remembered for figuring out how to continuously turn intelligence into deployed systems; he says the engineering reinvention to make building intelligent machines fast and iterative is the part almost nobody is building.