What AI-native looks like
Lee lets agents work overnight and starts the day with a personalised task summary. She says careful prompts help her avoid generic answers, and that her agent keeps learning while helping with everyday work.
Cordeira uses several agents and believes in an agentic user experience built around a single orchestrator. He uses AI at home and at work to speed up tasks, and describes himself as highly optimistic about AI.
For Lee, being AI-native also means adapting to a vocabulary and taxonomy that keep changing: new terms appear every two or three weeks. She connects the idea to Silicon Valley’s instinct to build, change the world and apply technology for the common good. Outside that culture, she said, AI still does not feel like a natural part of many people’s experience.
Cordeira’s test is more structural: imagine building an organisation from scratch with today’s AI capabilities. Simply adding AI to existing processes is not enough; the work itself needs to be reconsidered from end to end, with human involvement preserved.
The hard part is changing how people work
UBS has existed for 160 years, Cordeira said, and its established ways of working have produced financial results. That success makes change difficult: organisations scale what works, then build their technology, processes and staffing around it. Replacing those systems can mean dismantling them and starting again.
The obstacles are not only cultural. Lee pointed to legacy systems and resistance to data-centre construction and the infrastructure needed to expand AI. Leaders also face quarterly targets, while building something new may take more than one or two quarters. Cordeira added that finding data and tools can be difficult.
Lee described a startup inside Cisco, Energy Networking Systems, previously called the Engineering Sustainability Office. It has been operating for about four and a half years, with a team she described as “the size of two pizza boxes,” working through Cisco’s matrix structure.
Lee also argued that AI’s infrastructure has local costs that are easy to overlook. People depend on energy but often take it for granted; data centres can also affect water resources and local utility decisions. Cordeira, by contrast, said regulation slows progress but is fundamental to UBS, which takes its responsibilities to regulators seriously.
A product cannot change a habit by itself
Cordeira said UBS once assumed users would come if the company built a product for them. UBS developed advanced AI capabilities and promoted them widely, but financial advisers receive hundreds of emails a day and may have worked the same way for 30 years. The approach that worked best was personal help: going to advisers and sitting beside them as they learned to use AI.
That example puts a sharper edge on the AI-native claim. The constraint may not be whether a company can build an agent, but whether it can help people change routines that have served them well. I think this is the more revealing test in the discussion: adoption requires time and direct support, exactly what organisations under quarterly pressure may be least equipped to provide.
Lee expects AI to transform every job, including executive assistants’ work, as some already use agents. But she stressed that human assistants and small teams will still be needed for tasks requiring critical thinking. She described younger workers as quick to absorb information, while arguing that previous generations must help them develop the judgement their professions demand.
Cordeira’s longer-term picture is that people will see themselves more as process engineers than executors, managing agents that carry out processes. But the transition depends on more than young workers learning new tools. It asks established organisations to rebuild work without losing the human judgement that makes the work reliable.
Daily AI news
Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.
Only what matters — every day
Follow on X