Yuri Gurevich argues that algorithms are inherently deterministic and that probabilistic and quantum algorithms work by delegating non-algorithmic steps such as coin flips and quantum measurements to external oracles or assistants. He presents a theory of assisted algorithms with revised abstract state machine postulates, using a primality test and Grover's search as running examples.
Evan Kormos of Coinbase presents a practical framework for bringing AI agents into policy-driven, regulated workflows. He models processes as DAGs with loops inside each node, argues that a loop is only as trustworthy as its check, and shows how evals and traces create a new confidence mechanism. Agents then move through supervised, sampled, and exception-only stages.
This talk presents Aegon, a self-auditable key transparency scheme designed for global-scale encrypted messaging. It builds on Iron Dict's polynomial commitment approach to remove per-epoch global work, shards the dictionary for horizontal scaling, and uses proof caching to bound storage. The result is constant-size audit proofs that a lightweight client can verify itself over an entire history, rather than relying on external auditors.
Jedidiah Katz, longtime agent-harness engineer at Cursor and now at SpaceX, argues that fancy harnesses no longer add coding intelligence because models absorb those skills through a data flywheel. He contends the harness matters more than ever: it unlocks new frontiers like Grokbot and owns infrastructure, trust and security, and efficiency. He closes with predictions, including that 90% of agent work will run off the laptop within six months.
Jean-Denis Greze, Town's CEO and former Plaid CTO, explains why he shut down a nearly $1M-revenue AI tax business for SMBs and pivoted to Town, an AI assistant that works from email and calendar. The conversation with a16z's Alex Rampell covers founder decision-making under time pressure, product-market fit, the consumer-grade enterprise thesis, token allocation as a business model, neutrality versus frontier labs, horizontal versus vertical strategy, uneven AI diffusion, and how agents may reshape commerce and advertising.
Pamela Fox runs a guided workshop comparing GPT-5.5, Kimi K2.6, DeepSeek V4 Flash, and Mistral Large 3 on Microsoft Foundry, showing that benchmarks cannot predict how a model will behave in a specific application. The talk covers counting, RAG groundedness, vision, multilingual failures, tool calling, parallel tool calls in agents, string-check and LLM-as-a-judge evals, Microsoft's ASSERT tool for generating test sets, and DSPy prompt optimization that lets a cheap model satisfy a spec.