Inside the Personal AI Assistant Growing 10% a Day | Instinct Founder

Overview

• Speaker: Noah Shinn and Patrick O'Shaughnessy
• Channel: Invest Like The Best
• Main topic: Building Instinct, a personal AI assistant growing rapidly through unique interface design and trust networks.
• Purpose: To explore the product architecture, user psychology, growth dynamics, and strategic vision behind Instinct, a next-generation personal AI assistant. Noah Shinn, founder of Instinct, discusses building a personal AI assistant that operates through simple interfaces like text and voice, scaling rapidly at 10% day-over-day growth. The conversation covers the product's philosophy of acting as a proactive proxy for users, trust networks, economic models, and the future of human-computer interaction in an AI-native world.

Topic Map

Introduction to Instinct and Growth Dynamics

• Explanation: Introduction of Noah Shinn and Instinct, detailing its 10% day-over-day growth and positioning as a personal assistant.
• Key claims:
  • Instinct is growing at roughly 10% day-over-day.
  • The company focuses on building a personal assistant without a traditional app interface.
  • Growth is driven entirely by word-of-mouth through an invite-only model.
• Examples:
  • Users receive five invites to share with trusted contacts, leading to viral compounding loops.
• Terminology:
  • Instinct
  • personal assistant
  • invite-only
• Why it matters: Demonstrates rapid viral adoption of an AI consumer product without paid marketing spend.

Interface Philosophy and Everyday Intelligence

• Explanation: Discussion on why Instinct avoids a heavy app interface in favor of text, voice, and proactive agentic behavior.
• Key claims:
  • Instinct is not a new app or tool, but a new experience combining a phone and computer interface.
  • The AI acts proactively rather than just responding to prompts.
  • Simplifying the user experience reduces friction to zero.
• Examples:
  • Instinct calling a user to remind them of a pressing document deadline before 3 PM.
  • Scanning entire wardrobes and calendars to plan outfits and schedules autonomously.
• Terminology:
  • everyday intelligence
  • proactive behavior
  • zero friction
• Why it matters: Shifts software interaction from reactive tool usage to proactive, agent-driven assistance.

Trusted Person Networks and Social Dynamics

• Explanation: How Instinct coordinates between users through trusted social graphs while preserving privacy and safety.
• Key claims:
  • Users only connect with trusted people to share calendars, emails, and preferences.
  • Trust networks enable seamless multi-user scheduling and coordination.
  • Violating trust within the network has direct relationship implications.
• Examples:
  • Instinct coordinating a dinner party or ride-sharing across six friends by negotiating optimal schedules and routes.
• Terminology:
  • trusted person network
  • social dynamics
  • access control
• Why it matters: Extends AI utility from single-player productivity to multi-player social and professional coordination.

Business Models and Monetization

• Explanation: Examining the capital-intensive nature of personal AI assistants and potential monetization strategies.
• Key claims:
  • Traditional consumer internet businesses rely on ads, which incentivize capturing user attention.
  • Instinct aligns with user intent rather than maximizing screen time.
  • Capital intensity requires careful consideration of compute costs and unit economics.
• Examples:
  • Comparing take-rates in travel and restaurant booking against traditional aggregators like Expedia or OpenTable.
• Terminology:
  • take-rate
  • unit economics
  • compute inference cost
• Why it matters: Highlights the economic challenges and alignment differences between AI agents and ad-supported platforms.

Security, Trust, and the Future of Software

• Explanation: Addressing the challenges of data privacy, safety, and the eventual collapse of traditional software interfaces.
• Key claims:
  • Building trust takes weeks, but retention remains high (80% retention when sensitive info is shared).
  • AI agents will eventually replace legacy software applications as the primary interface.
  • Proactivity requires rigorous safety firewalls and oversight.
• Examples:
  • 40% of users share personal credit card information with Instinct within three weeks.
• Terminology:
  • retention rate
  • firewall
  • inference cost
• Why it matters: Establishes trust as the core moat and technical bottleneck for autonomous personal agents.

Key Points

Proactive vs. Reactive AI

• Explanation: True personal assistants must anticipate needs and act proactively without requiring constant prompting from the user.
• Evidence: Instinct calling users with urgent deadlines or booking travel itineraries automatically based on inferred preferences.
• Practical implication: Product design must shift from chatbot prompt interfaces to ambient, autonomous background execution.

Trust as a Compounding Moat

• Explanation: Sharing sensitive data like credit cards and calendars builds high retention and trust over a multi-week period.
• Evidence: 80% retention rate for users who share at least one sensitive piece of information with Instinct.
• Practical implication: Early product onboarding must prioritize security, data control, and gradual trust-building.

Multi-Player AI Coordination

• Explanation: AI agents communicating with other users' AI agents eliminates friction in scheduling, planning, and commerce.
• Evidence: Coordinating rideshares, dinner bookings, and meetings across groups of friends automatically.
• Practical implication: Network effects emerge when AI agents transact and coordinate on behalf of multiple users.

Frameworks, Models & Processes

Trust-Based Scaling Framework

• How it works: Gradually expanding AI autonomy as the user shares sensitive data and establishes trust over time.
• Components:
  • Frictionless onboarding
  • Granular access controls
  • Proactive background execution
  • High retention through personal context
• When to use: Building personal AI agents that handle sensitive financial, scheduling, and personal data.

Examples & Case Studies

An Instinct user needed to coordinate a ride for six friends after a dinner party.

• Illustrates: Multi-player AI coordination and route optimization.
• Lesson: AI agents interacting with each other reduce social coordination friction to near zero.

Instinct users scanning their entire wardrobes and bodies to plan weekly outfits.

• Illustrates: High-intent consumer use cases for proactive AI agents.
• Lesson: Consumers willingly share sensitive data when the utility and time savings are extreme.

Actionable Takeaways

• Immediate:
  • Focus on understandability and predictability rather than raw capability in AI agent design.
  • Design invite-only distribution loops to manage compute scaling and viral growth.
• Strategic:
  • The future of software is agent-native, where traditional apps disappear behind conversational and proactive interfaces.
  • Alignment with user intent creates a superior business model compared to ad-driven attention economies.
• Questions to investigate:
  • How will compute inference costs scale relative to user growth for autonomous agents?
  • What new security paradigms are required as AI agents gain full access to personal financial accounts?

Claims Worth Verifying

• Instinct is growing at 10% day-over-day. (metric)
• 40% of users share personal credit card information within three weeks. (metric)
• There is an 80% retention rate for users who share sensitive information. (metric)

Notable Quotes

"This is probably the most exciting software race ever." (at 0:31) "We're not focused on capability, we're only focused on understandability." (at 82:15)

Compressed Summary

• Instinct is a personal AI assistant growing at 10% day-over-day via invite-only loops.
• The product emphasizes proactivity and zero-friction interfaces over traditional app menus.
• Trust and data security are the primary moats, driving an 80% retention rate among active users.
• Multi-player AI coordination allows agents to interact with other agents to solve complex scheduling tasks.
• Keywords: artificial intelligence, personal assistant, agentic workflow, trust network, software interface
• Core insight: The future of software is shifting from reactive applications to proactive, trusted personal AI agents that coordinate seamlessly across human networks.