Frontier AI Architect
by Amar
You Can't Verify What the Weights Learned: AI Model Supply Chains
About this podcast
Architectural blueprints, infrastructure economics, and strategic moats at the frontier of artificial intelligence.
High-signal deep dives for founders, tech executives, and systems engineers. Exploring how to move beyond thin wrappers, architect autonomous enterprise agents, and capture the $10T labor market.
Read the companion monographs and subscribe: https://frontieraiarchitect.substack.com
Direct Inquiries & Briefings: [email protected]
Episodes
- You Can't Verify What the Weights Learned: AI Model Supply Chains · 20 minThis episode's audio is AI-generated, based on an article written by David Amar. A Hugging Face malware campaign passed every scan while its safetensors file held 42 bytes. This…
- AI Diligence: Check Logs, Not Demos (What Nate and Presto Teach) · 21 minThis episode's audio is AI-generated, based on an article written by David Amar. An investor asked the right question about Nate's AI and accepted an email as the answer. This…
- Nobody Needs to Steal Your Weights: How Models Really Get Copied · 21 minThis episode's audio is AI-generated, based on an article written by David Amar. Three US agencies, Anthropic, Google and OpenAI all describe the same thing in 2026: models…
- Your AI Supplier Can Drop You in 5 Days: Capacity, Not Price · 20 minThis episode's audio is AI-generated, based on an article written by David Amar. Windsurf offered to pay in full and still lost nearly all of its direct Claude capacity with…
- Agents Can Pay. Who Pays When They Get It Wrong? · 18 minThis episode's audio is AI-generated, based on an article written by David Amar. Agents can already pay on cards and on stablecoins, but the measured volume is small and no rule…
- How Long Should Your Model Think? The Real Cost of Reasoning Effort · 22 minThis episode's audio is AI-generated, based on an article written by David Amar. OpenAI and Anthropic both got the reasoning setting wrong for their own flagship products, in…
- Amazon v. Perplexity: Who Is the User When an Agent Shops? · 20 minThis episode's audio is AI-generated, based on an article written by David Amar. An appeals court said the shopper, not Perplexity's agent, is the one accessing Amazon, and…
- The Death of Volume Outbound: Sales Agents and AI Answers · 21 minThis episode's audio is AI-generated, based on an article written by David Amar. The first AI sales agents automated the one part of outbound that was already failing: sending…
- Solo Unicorns: What a Two-Person, $401 Million Company Proves · 21 minThis episode's audio is AI-generated, based on an article written by David Amar. A two-person telehealth company reported $401 million in first-year sales, along with an FDA…
- Compute Burn vs. Dilution: Who Really Pays for Your GPUs · 22 minThis episode's audio is AI-generated, based on an article written by David Amar. Stability AI's founder was still the majority shareholder when he left a company whose compute…
- The EU AI Act Deadline Moved. The Engineering Work Did Not. · 19 minThis episode's audio is AI-generated, based on an article written by David Amar. Days before the EU AI Act's high-risk rules were due to apply, the EU pushed them back by up to…
- The Inference Bill: Cloud APIs, Your Own GPUs and the True Cost · 19 minFor a fixed level of quality, token prices fall about tenfold a year, and AI bills keep rising. This episode explains why, what running your own GPUs really costs, and how to…
- Beyond Naive RAG: Why Retrieval Is Not Memory · 20 minLegal research tools built on retrieval and on some of the best document collections in any industry still hallucinated on 17% to 33% of test queries. This episode explains where…
- MCP Tool Poisoning: Your Agent Trusts Every Server It Meets · 18 minA fake Postmark package copied every email its users sent through an AI assistant. This episode explains why the protocol that connects agents to tools cannot enforce trust, and…
- Who Is Your Agent? Identity for Software That Acts for You · 14 minStolen OAuth tokens from an AI chatbot opened hundreds of Salesforce tenants, and every call looked legitimate. This episode explains how to give agents an identity of their own…
- Evals-Driven Development: Your Test Suite Is the Moat · 21 minOne instruction in a system prompt measurably degraded a coding product at the company that built the model. This episode explains how to build the test suite that catches that…
- The Confused Deputy: Why Filters Lose to Prompt Injection · 20 minA single email made Microsoft 365 Copilot leak data, past three layers of filtering. This episode argues that agent security is about what authority the model's output carries…
- Per-Seat Pricing Is Not Dead: What to Charge When AI Does the Work · 21 minSalesforce priced its AI agents three different ways in about a year and ended up close to a seat again. This episode explains what that teaches anyone who has to put a price on…
- Services-as-Software: Selling the Work, Not the Tool · 19 minKlarna claimed its AI assistant did the work of 700 agents, then went back to hiring people, then claimed even bigger savings. This episode uses that story to explain the shift…
- Beyond the Wrapper: Why the Model Is Not Your Moat · 13 min · Season 1 Episode 1In October 2022 Jasper raised $125M at a $1.5B valuation. About six weeks later its own supplier released ChatGPT for free. This episode is a teardown of the wrapper trap: why a…








