GPT-5

Inside AI’s $10B+ Capital Flywheel — Martin Casado & Sarah Wang of a16z

Feb 19, 2026 · 55:31

Martin Casado and Sarah Wang of a16z argue that AI’s capital flywheel—where model labs translate funding directly into capability gains and revenue growth in weeks—is creating a new financing playbook that blends venture and growth, with rounds acting as compute contracts. They warn that frontier labs like Anthropic can potentially raise more money than the entire app ecosystem built on their APIs, allowing them to outspend and consume those layers. The episode examines the AGI vs. product dilemma in GPU allocation, the war for talent where $10M+ packages break early-stage founder math, and Cursor as a case study of building up from the app layer while training down into its own models. They also identify “boring” enterprise software as the most underinvested opportunity and note that robotics lacks a ChatGPT moment that would justify current funding levels.

[State of Post-Training] From GPT-4.1 to 5.1: RLVR, Agent & Token Efficiency — Josh McGrath, OpenAI

Dec 31, 2025 · 27:34

Josh McGrath, an OpenAI post-training researcher, states that real post-training innovation lies in data quality and signal trust, not optimization methods, with RLVR and token efficiency central. He moved from pre-training (3% compute gains) to post-training (40% behavior change), describes the infrastructure chaos of RL runs, and cites GRPO from DeepSeek Math as underappreciated for providing verifiable reward signals. GPT-5 to GPT-5.1 bumped evals while slashing tokens, emphasizing token efficiency over wall-clock time. The shopping model features interruptibility and chain-of-thought transparency, and personality toggles (Anton vs Clippy) are a key differentiator. For long context, he argues agents with graph walks may be more important than 10M-token windows. He concludes that the education system fails to produce people skilled in both distributed systems and ML research, a critical combination as bottlenecks shift.

[State of RL/Reasoning] IMO/IOI Gold, OpenAI o3/GPT-5, and Cursor Composer — Ashvin Nair, Cursor

Dec 30, 2025 · 45:13

Ashvin Nair, now ML lead at Cursor, traces his path from Berkeley robotics and an OpenAI Dota-era internship to OpenAI's reasoning team (which grew from a dozen to 300+ people) and explains why IOI Gold in 2022 felt like solving AI but didn't change the world—because RL doesn't generalize beyond training distribution. He argues most RL research from 2017-2022 overfit to benchmarks, rewarding complex ideas over simple ones that scale. At Cursor, he sees a unique opportunity for continual learning with policy updates every two hours and product-model co-design, keeping engineers in the loop instead of context-switching. His bet is that the next paradigm shift is continual learning with infinite memory: models experience something once and never forget it, storing millions of deployment tokens in weights without overloading capacity.

⚡️GPT5-Codex-Max: Training Agents with Personality, Tools & Trust — Brian Fioca + Bill Chen, OpenAI

Dec 26, 2025 · 27:46

OpenAI’s Brian Fioca and Bill Chen discuss the launch of Codex Max, a long-running coding agent designed to work 24+ hours, manage its own context, and spawn sub-agents. They reveal how GPT-5 training prioritized personality traits—communication, planning, and self-checking—to build developer trust, and how Codex develops tool habits like preferring 'rg' over 'grep'. The abstraction layer is shifting from models to full-stack agents, enabling partners to plug Codex into tools like VS Code or Zed. They highlight applied evals that measure real-world impact over academic benchmarks, and the need for multi-turn eval APIs. Their 2026 vision: coding agents trusted enough to handle the hardest refactors at any company, not just top-tier firms, and general enough to automate personal workflows like organizing desktops or email.

How Zyphra went all-in on AMD + Why Devs feel faster with AI but are slower — with Quentin Anthony

Nov 3, 2025 · 1:00:02

Quentin Anthony, head of model training at Zyphra and advisor at EleutherAI, explains why his company moved all training to AMD MI300X GPUs, outperforming Nvidia H100s on certain workloads thanks to 192GB VRAM and higher memory bandwidth, and describes his kernel development approach of writing directly in ROCm or GPU assembly rather than using Triton. He also shares his experience in the METR study on AI coding productivity, where he was one of the few developers with measurable speedup, and offers tips: timebox AI use, avoid the slot machine effect, maintain context hygiene, and use direct API over tools like Cursor. Additionally, he argues that open source AI research benefits from siloed focused teams with guaranteed funding over grand collaborations, and notes that kernel datasets alone won't solve GPU programming due to evaluation challenges.

Terminal-Bench: Pushing Claude Code, OpenAI Codex, Factory Droid, et al to the limits

Oct 18, 2025 · 35:27

Alex Shaw and Mike Merrill, creators of Terminal-Bench, explain how their coding agent benchmark became an industry standard adopted by Anthropic, OpenAI, and leading agent companies. They bet on terminal-based interaction over GUI because text is the modality that works best with models, and designed tasks as containerized environments with instructions and test scripts. Nicholas Carlini at Anthropic became an early champion, leading to Terminal-Bench being featured on Claude’s model card without notice. They built the minimal Terminus agent to isolate model capabilities from agent optimizations, and their framework adapts existing benchmarks like SWE-bench. Future plans include cloud hosting, multi-dimensional evaluation incorporating cost and economic value, and enabling RL post-training on benchmark tasks.

DevDay 2025: Apps SDK, Agent Kit, MCP, Codex and why Prompting is More Important than Ever

Oct 7, 2025 · 44:52

Sherwin Wu and Christina Huang from OpenAI's platform team discuss the new AgentKit suite and Apps SDK, arguing that visual agent builders and prompt optimization remain critical for production AI. They detail AgentKit's components—Agent SDK, visual Builder, Evals, and Tool Registry—and how Apps SDK inverts the website-chatbot paradigm by embedding apps inside ChatGPT. OpenAI adopted Anthropic's MCP protocol for tool connectivity, and the team dogfoods these tools for their own customer support at openai.com. They explain that automated prompt optimization and evaluation pipelines are essential for scaling agents, and share how Codex is used internally for code generation and PR reviews. The episode also covers the Service Health Dashboard for real-time API SLO tracking, reflecting OpenAI's reliability investments.

⚡️Traversal: Causal ML and Reinforcement Learning

Oct 5, 2025 · 45:06

Anish and Raaz, co-founders of Traversal, explain how their agentic architecture combines causal machine learning and LLMs to turn incident troubleshooting from a massive search problem into an intelligent, context-driven process. Both have PhDs in causal ML and RL from MIT and Berkeley, and they applied that expertise to root cause analysis in complex enterprise systems with petabytes of fragmented data across logs, metrics, traces, and Slack. Their system dynamically combines statistical tests (for time-series) with semantic understanding (via LLMs) to winnow billions of signals to root cause candidates in under two minutes. They note that reasoning models like o3 are critical for complex incidents, while Claude is better for tool calling and unsticking. Traversal uses a mix of infrastructure size and investigation count for pricing, and sees self-healing as a continuum: 10–20% of issues can be autonomously resolved now, with 30–40% requiring senior engineer sanity-check within 6–12 months. They are hiring in New York and emphasize that building proprietary evals is core IP, making public benchmarks a tension.

Context Engineering for Agents - Lance Martin, LangChain

Sep 11, 2025 · 1:03:18

Lance Martin from LangChain discusses context engineering for AI agents, arguing that managing context from tool calls is the key challenge and introducing five techniques: offloading, reducing, retrieval, caching, and context isolation. Offloading saves raw tool call outputs to disk and passes only summaries, drastically cutting token usage—Martin's deep research agent went from 500,000 tokens per run to much less. Manus uses file system offloading and warns that irreversible pruning risks information loss, while Cognition advocates fine-tuned summarization for agent-agent boundaries. For retrieval, Martin contrasts Windsurf's multi-step indexing with Claude Code's agentic search using grep and llms.txt, finding the latter more effective. He applies the Bitter Lesson to agent engineering: his own Open Deep Research evolved from a highly structured workflow to a simple agent loop as models improved, and he warns against rigid abstractions that hinder adaptation.

Greg Brockman on OpenAI's Road to AGI

Aug 15, 2025 · 1:08:37

Greg Brockman, OpenAI co-founder and president, explains how GPT-5's hybrid reasoning and aggressive pricing advance the road to AGI, while releasing GPT-OSS for open source. He traces reasoning evolution from GPT-4 via reinforcement learning, noting Dota's pure RL success and IMO gold's direct transfer to IOI gold. Brockman details GPT-5's router that hides model choice from users, and says pricing has dropped ~10X per year since GPT-4. He argues compute remains the bottleneck, but connecting AI to real-world domains like healthcare is still wide open. Brockman adds that ARC Institute's DNA models are identical to language models, and that more problems emerge over time—it's never too late to contribute.

⚡️ARC-AGI-3: The Interactive Reasoning Benchmark

Jul 18, 2025 · 39:41

Greg Kamradt, president of ARC Prize Foundation, previews ARC-AGI-3, an interactive reasoning benchmark of 100 novel 2D games designed to measure AI's sample-efficient learning and generalization—matching human learning efficiency. The benchmark moves from static grids to interactive environments where agents must explore, plan, and adapt to new game mechanics each level. Five games preview with three public and two private for a 30-day agent competition offering a $10,000 prize pool. Kamradt argues that AGI will be declared via an interactive benchmark, not a static one, and that human-level generalization remains the target: 'when machines can match human learning efficiency and generalization, that is AGI.' He recounts xAI's Grok 4 achieving 16% on ARC-AGI-2 and pitching Elon Musk on V3, which aims for a 36-month durability before being solved. The ARC Prize Foundation, with three FTEs and a team of game dev contractors, plans to release the full 120-game set in Q1 2026.