THE AI INTERVIEW
“ Every stage of my career has really been about understanding increasingly complex systems”
Shayan Mohanty Chief Data & AI Officer Thoughtworks
“ Traditional systems are deterministic, so you input X and reliably get Y. Generative AI hallucinates, makes bizarre contextual mistakes and can fail in ways that are incredibly difficult to predict ahead of time.”
Even once AI produces a meaningful share of a company’ s output, Shayan argues that isolated automation simply does not suffice and, instead, entire workflows need rebuilding around the possibility of error.
“ The organisations succeeding right now are redesigning systems around verification, supervision and recovery, not around the assumption the AI will always be correct,” he adds.“ They are building operational safety nets into the workflow itself.”
Judgement over craftsmanship The aforementioned shift explains why Shayan believes human judgment is actually becoming more valuable as AI adoption grows in the workplace.
“ We are leaving an era of pure craftsmanship, where your value was measured by how fast you could physically execute a task,” he explains.“ We’ re entering an era of structural design.”
With AI able to generate output at breakneck speed, Shayan says the most in-demand human skills are direction and delegation:“ You need strong human judgment to set the boundaries, understand the business context and know how to safely steer a fleet of digital tools.”
Judgment itself is difficult to hire for, Shayan emphasises, noting that it depends heavily on experience and context rather than raw aptitude. Thoughtworks’ s approach is to change each engineer’ s relationship with the code.
24 August 2026