Why steady expertise still decides who gains from workplace automation

The question of whether intelligent software will take our jobs carries real anxiety, yet Oxford professor Pinar Ozcan answers it with calm historical framing . She places the current wave in a longer sequence, where each major technology first frightens workers and later reshapes tasks. Her point is not comfort for its own sake but a call to read the pattern carefully. What matters, she suggests, is who learns to work alongside the new systems and who is left without access to training.

Ozcan was born in Istanbul as the only daughter of two entrepreneurs. When she was six, her parents founded what she describes as one of Turkey's first leadership training firms. She grew up around weekend seminars, copying machines, envelopes, and visiting speakers. Entrepreneurship fascinated her, but she also saw its daily strain up close. That double impression later pulled her toward studying entrepreneurship as an academic rather than living with its constant pressure.

From scholarship years to Oxford leadership

She went to Stanford on a scholarship for a master's degree and then a doctorate. She arrived as the dot-com boom accelerated and stayed through its sharp collapse at the turn of the century. Watching valuations soar and then fall taught her a durable lesson. Technology does not merely add new products; it rewrites how firms compete. That insight shaped her later research on platforms, finance, and digital change.

Ozcan is now Professor of Entrepreneurship and Innovation at the Said Business School, University of Oxford. She founded and directs the Oxford Future of Finance and Technology initiative and serves as academic director of the Oxford Entrepreneurship Centre. Through that centre she supports women founders through a global founders program. She has advised the CMA, the FCA, the European Commission, the OECD, and the Bank of England, and she holds a Stanford doctorate in Management Science and Engineering. The profile records published on PinarOzcan confirm this role description.

She says rooms in her field often looked balanced between men and women, which she welcomed. Yet she also sensed that some panel invitations were driven by optics, suspecting she was added so the lineup would not look all male. Rather than refuse those stages, she accepted them and tried to make her remarks open doors for others. The pattern pushed her research toward women founders, who remain a small share of entrepreneurs and receive a far smaller share of funding. She argues that only deeper structural steps can shift those ratios.

Classrooms, banks, and new job paths

Her team brings founder stories into secondary schools, teaching both girls and boys through examples of women who built companies. The work has reached a school in Turkey, a school in Scotland, and a wider international school network, with a public webinar series planned next. The goal is to normalize the idea that building a company is a learnable path rather than a rare gift. A randomized evaluation of a fifty-hour entrepreneurship program for ninth graders in India, documented in Worldbank open knowledge records, gives this kind of school-level effort a serious evidence base.

Banks offer a clear preview of job change, as branch closures move routine work to online and phone channels. Ozcan notes that affected staff were often retrained for customer support roles rather than dismissed outright. From this she draws a simple formula of deep craft plus automation fluency , where secure work belongs to people who combine domain judgment with the ability to direct machine systems. Those without retraining face the sharpest displacement risk. A January 2026 IMF staff discussion note documents higher wages in vacancies demanding AI skills, underscoring the pay gap between those who adapt and those who cannot.

Smaller firms, wider ecosystems

The same logic reaches office services, where one skilled operator guiding capable software agents can replace the output of a small bureau on its own . Ozcan expects firms to become smaller on average while growing more numerous, linked through partnerships rather than large payrolls. In that world the ecosystem idea matters more than headcount. Early marketplace examples described by Forbes illustrate how agent toolkits let modest businesses serve larger clients.

Tiny product companies with single-digit staff show what this can look like in practice. Ozcan points to compact teams behind products such as Base44, AI Apply, and Oleve as signs of how far small groups can reach with machine help. The design logic runs from modular tasks to agent autonomy, then to smooth integration under steady human supervision . Managing people and software agents as one coordinated crew becomes the core skill. This micro-firm typology with single-digit teams is analyzed by Researchleap across four cases. The blended human-agent management frame is set out by Birlasoft in its 2025 study.

Health care may show the most hopeful use, by spotting faint early risk signals five or six years before illness takes hold. Ozcan also describes training physicians in distant regions with machine guidance so simpler procedures can happen locally, citing as an example relayed by the speaker a program linked to hundreds of thousands of operations. That account should be read as an unverified illustration rather than an established result. She adds that hospitals increasingly work alongside machine-intelligence firms to build these tools. Work on a generative model predicting the onset of more than 1,200 diseases from medical records, summarized by Nature in its September 2025 news report, shows the research direction behind such hopes.