Dwarkesh Patel

Interviewer and host of the Dwarkesh Patel podcast. In [src-042], he runs a blackboard-style technical interview with Reiner Pope on how frontier models are trained and served.

Key facts

  • Type: Podcast host / technical interviewer
  • Relevant source: "How GPT, Claude, and Gemini are actually trained and served"
  • Format in this source: Long-form blackboard lecture focused on model architecture, ML infrastructure, inference economics, and training systems [src-042]
  • Why it matters: The interview connects user-visible API behavior, such as fast modes, context pricing, and cache pricing, to low-level hardware constraints [src-042]

Related entities

Related concepts

Source references

  • [src-042] Dwarkesh Patel — "How GPT, Claude, and Gemini are actually trained and served – Reiner Pope" (2026-04-29)

Robin Cartier perspective

This page is part of Robin Cartier's working AI knowledge graph: a practical research layer for production AI, recommendation systems, experimentation, GEO, and agentic web readiness.

The useful next step is to connect this concept back to applied product leadership and operating models.

Recommended next

Keep reading from this thread

From 500 indexed pages and articles.

  1. Wiki concept Roofline Analysis for LLM Serving Back-of-the-envelope method for estimating LLM serving latency by comparing compute time against memory-fetch time. Related by 042
  2. Wiki concept KV Cache Tiering Serving strategy for deciding whether to keep KV-cache state in fast memory, move it to slower storage, or delete and rematerialize it later Related by 042
  3. Wiki concept LLM Serving Batching Serving optimization where multiple user sequences are decoded together so the cost of reading model weights is shared across many generated tokens. Related by 042