Unlike traditional deterministic products, AI models are probabilistic; the same query can yield different results. This uncertainty requires designers, PMs, and engineers to align on flexible expectations rather than fixed workflows, fundamentally changing the nature of collaboration.

For 50 years, software development operated at one level of abstraction: humans writing software. AI agents introduced a new level: humans telling models to write software. Now, advanced routines introduce a third level: models telling other models what to do, representing an unprecedented acceleration in programming abstraction.

Traditional software relies on predictable, deterministic functions. AI agents introduce a new paradigm of "stochastic subroutines," where correctness and logic are abdicated. This means developers must design systems that can achieve reliable outcomes despite the non-deterministic paths the AI might take to get there.

Leading engineers like OpenAI's Andre Karpathy describe recent AI tools not as incremental improvements but as the biggest workflow change in decades. The paradigm has shifted from humans writing code with AI help to AI writing code with human guidance.

Researchers understood the principles of building rational AI systems with uncertainty decades ago. Their adoption was stalled not by theoretical weakness but by intractable computational demands. With modern hardware, these once-abandoned ideas are finally becoming feasible.

Unlike other industries accustomed to deterministic software, the finance world is already familiar with non-deterministic systems through stochastic pricing models and market analysis. This cultural familiarity gives financial professionals a head start in embracing the probabilistic nature of modern AI tools.

Future literacy requires understanding concepts beyond deterministic algorithms. As AI tools become more prevalent, users will need to grasp probabilistic and stochastic systems to effectively build with and manage them, recognizing that outputs are not always perfectly reproducible.

Unlike past software that performed deterministic tasks, future software will be infused with AI, giving it an 'opinion' and proactive intelligence. Nikesh Arora predicts this fundamental shift means the entire software industry will be rewritten over the next 10 years, creating massive disruption and opportunity.

Instead of writing code and then tests, developers could define precise mathematical properties. AI would then synthesize both the program and a formal proof that the program meets those specifications, ensuring correctness by design and flipping the current code-first paradigm.