Interview: Project Tbd Download W Max August — 2026-04-27
Key Themes
Origin Story & Mission Alignment: Max, Dustin, and Bliss traced the project’s lineage from first-principles geopolitical reasoning (multipolar world → economic outcomes → technological outcomes) to a specific semiconductor supply chain thesis. The intellectual honesty about how long it took to get to conviction (~5 hours of cumulative whiteboarding) was notable. Max was brought in as an external sounding board, not a co-founder.
Digital Twin via Ontology: Bliss’s Palantir background is doing real work here. The ontology framing — a human-readable map of supply chain data objects rather than a pile of spreadsheets — is both the product vision and the moat hypothesis. The connection to LLM-powered simulation workflows is well-reasoned and maps to Palantir’s own public positioning in the AI era.
Compliance Wedge as Data Acquisition Strategy: The core go-to-market logic is sound: use regulatory urgency (UFLPA, CHIPS Act, ownership restrictions, Applied Materials/Cadence-scale enforcement) to create a workflow companies will pay for and contribute proprietary data into. Defensive compliance (required filings) and offensive compliance (competitive positioning for government contracts) as two flavors of the same entry point.
Flywheel Model: Every new customer reduces verification burden for overlapping suppliers — a genuine network effect mechanism, not just an analogy. This is the structural moat argument, but it hasn’t been stress-tested against real procurement workflows.
Grand Slam Opportunities Remain Unvalidated: Insurance, trading/benchmarks, government intelligence, marketplace, prediction markets — all identified but none sized with real numbers or validated with a single customer conversation. The Gemini and Claude memos provide useful external framing, but internal customer discovery is at zero.
Notable Quotes
- “Semis are the oil of the 21st century — whichever country controls the supply chain controls the dominant geopolitical power going forward.” (Steve Blank, paraphrased by Dustin)
- “An ontology is a word Palantir people bludgeon around like a club and use to solve any problem, but can’t explain what it actually means.” (Bliss)
- “We’ve pushed the AI tools really hard… but it’s been very heads down, non-interactive.” (Dustin, flagging the external research gap)
- “He’s trying to build a physical twin and we’re trying to build a digital twin.” (Dustin, on the LaPolice comparison)
Surprises
- Zero external customer conversations to date. For a team this analytically sharp, the absence of any semiconductor industry interviews is the single biggest process gap. All conviction is AI-assisted and inductive.
- Max was not deeply briefed. He came in with ~30 minutes of Dustin context. This was a true cold download, not a structured advisory session.
- The Palantir ontology framing may be both an asset and a liability. It’s a genuine architectural advantage, but it could anchor the team to an enterprise data model that’s hard to wedge into without significant integration lift.
Open Questions
- Will mid-tier automotive/electronics buyers (not TSMC-scale) actually pay for compliance tooling, or will they rely on consultants?
- What does the competitive landscape look like for UFLPA compliance specifically — who already has traction?
- How does the flywheel survive the cold-start problem if major suppliers refuse to participate?
- What is the enforcement risk for the compliance wedge if regulatory priorities shift under the current administration?
- Does Max have warm introductions into insurance or government contacts as discussed?