Open‑source math: NOMOS‑1 Putnam results

They opened energized and focused almost immediately on GPT 5.2, why the benchmarks matter less than behavior, and what actually feels different when you build with it. Brian shared that he spent four straight hours rebuilding his internal gem builder using GPT 5.2, specifically to test whether OpenAI finally moved past brittle master and router prompting. The rest of the episode mixed deep hands on prompting work, real world agent behavior, smaller but meaningful AI breakthroughs in vision restoration and open source math reasoning, and reflections on where agentic systems are clearly heading.

Key Points Discussed

GPT 5.2 shows a real shift toward higher level goal driven prompting

Benchmarks matter less than whether custom GPTs are easier to build and maintain

GPT 5.2 Pro enables collapsing complex multi prompt systems into single meta prompts

Cookbook guidance is critical for understanding how 5.2 behaves differently from 5.1

Brian rebuilt his gem builder using fewer documents and far less prompt scaffolding

Structured phase based prompting works reliably without master router logic

Stress testing and red teaming can now be handled inside a single build flow

Spreadsheet reasoning and chart interpretation show meaningful improvement

Image generation still lags Gemini for comics and precise text placement

OpenAI hints at a smaller Shipmas style release coming next week

Topaz Labs wins an Emmy for AI powered image and video restoration

Science Corp raises 260M for a grain sized retinal implant restoring vision

Open source Nomos One scores near elite human levels on the Putnam math competition

Advanced orchestration beats raw model scale in some reasoning tasks

Agentic systems now behave more like pseudocode than chat interfaces

Timestamps and Topics

00:00:00 👋 Opening, GPT 5.2 focus, community callout

00:04:30 🧠 Initial reactions to GPT 5.2 Pro and benchmarks

00:09:30 📊 Spreadsheet reasoning and financial model improvements

00:14:40 ⏱️ Timeouts, latency tradeoffs, and cost considerations

00:18:20 📚 GPT 5.2 prompting cookbook walkthrough

00:24:00 🧩 Rebuilding the gem builder without master router prompts

00:31:40 🔒 Phase locking, guided workflows, and agent like behavior

00:38:20 🧪 Stress testing prompts inside the build process

00:44:10 🧾 Live demo of new client research and prep GPT

00:52:00 🖼️ Image generation test results versus Gemini

00:56:30 🏆 Topaz Labs wins Emmy for restoration tech

01:00:40 👁️ Retinal implant restores vision using AI and BCI

01:05:20 🧮 Nomos One open source model dominates math benchmarks

01:11:30 🤖 Agentic behavior as pseudocode and PRD driven execution

01:18:30 🎄 Shipmas speculation and next week expectations

01:22:40 🏁 Week wrap up and community reminders

The Daily AI Show Co Hosts: Brian Maucere, Beth Lyons, and Andy Halliday