Jason Reader—my former COO at Remington and now COO of Davidson Hospitality—joined me for a candid, wide-ranging conversation that pulls no punches.
We cover:
Why Jason ultimately decided to leave Remington after my departure in Q2 2025
What it’s really like stepping into a COO role in the middle of one of the hardest budget seasons in recent memory
Why third-party hotel management is more commoditized than ever—and how great operators still differentiate
Jason’s leadership philosophy: hungry, humble, people-first
AI, Power, and the Future of Hospitality
I also want to share a piece of research that’s been shaping how I think about AI written by the preeminent minds in AI: Eric Schmidt (Former Google CEO); Alexandr Wang (Head of AI at Meta); Dan Hendrycks (prominent AI safety expert).
The white paper, Superintelligence Strategy, was written from a national-security lens in 2025, but its implications for business—and hospitality in particular—are profound. I’ve provided a link to the full publication which is 41 pages in total.
📄 Download the Full white paper here:
https://drive.google.com/file/d/1JVPc3ObMP1L2a53T5LA1xxKXM6DAwEiC/view?ref=maginative.com
Here are my thoughts on “What should Hospitality take from this AI literature?”
1. Establish AI Governance Before Scaling AI
Why it matters:
The paper warns that loss of control happens gradually—not through one big mistake, but through many “reasonable” automation decisions.
Actions:
Create an AI Steering Committee (Ops, IT, Finance, Legal).
Define where AI is allowed to decide autonomously vs. where humans must approve (pricing floors, comping, labor cuts).
Require human-in-the-loop controls for revenue, procurement, and guest-facing decisions.
Maintain the ability to override or shut down AI systems quickly.
Hospitality example:
AI can recommend labor cuts—but a human must approve reductions that impact service scores or brand standards.
2. Treat Your Data Like a Strategic Asset (Not a Vendor Input)
Why it matters:
The paper emphasizes information security as a first-order risk. In hospitality, data leakage = lost advantage.
Actions:
Inventory all data flowing into AI tools (PMS, POS, CRM, RevPAR history, payroll).
Ensure contracts prohibit vendors from training models on your data.
Segregate owner-level data, brand data, and property-level data.
Use private or tenant-isolated AI environments for sensitive workflows.
Hospitality example:
Your pricing logic, labor optimization models, and procurement benchmarks should not become your competitor’s advantage via shared AI platforms.
3. Avoid “Black Box” AI in Core Operations
Why it matters:
The white paper stresses that systems operating beyond human understanding increase catastrophic risk.
Actions:
Prefer AI systems that explain recommendations (why rates moved, why labor changed).
Require audit logs for:
Pricing changes
Vendor selections
Marketing spend reallocations
Test AI outputs against brand standards and service metrics, not just margins.
Hospitality example:
If RevPAR improves but guest satisfaction collapses, the AI failed—even if EBITDA rises short-term.
4. Build AI Redundancy & Vendor Optionality
Why it matters:
The paper warns against over-centralization and dependency—once systems are embedded, they’re hard to unwind.
Actions:
Never allow one AI vendor to control:
Revenue + labor + procurement simultaneously
Maintain manual fallback playbooks for:
Rate setting
Labor scheduling
Vendor ordering
Avoid proprietary lock-ins that prevent switching providers.
Hospitality example:
An AI-driven procurement platform should never be the only way you can place emergency orders.
5. Use AI Aggressively Where Risk Is Asymmetric
Why it matters:
The paper differentiates between low-risk and high-risk AI domains. Hoteliers should do the same.
Best near-term AI use cases:
Demand forecasting
Dynamic pricing recommendations (with guardrails)
Procurement benchmarking
Invoice auditing & fraud detection
Predictive maintenance
Marketing attribution & personalization
Delay or tightly govern:
Automated comping/refunds
AI-driven workforce reductions without human review
6. Train Leaders, Not Just Teams
Why it matters:
AI risk is managerial, not technical. The paper emphasizes decision-making quality at the top.
Actions:
Require GM, Regional Ops, and Asset Managers to:
Understand AI limitations
Spot hallucinations or incentive misalignment
Run AI “fire drills”:
“What happens if this system fails at 6pm on a sold-out night?”
Cheers to 2026 being the Best Yet!!
-Sloan