The $10B AI Assistant Challenging Meta's Muse, Grok's Bots, and OpenAI's Dots
The guest sees personal agents as a potentially larger software opportunity than code generation and describes how they could reshape reservations, travel, and online ordering. He also discusses monetization trade-offs, testing and trust, and Instinct's rapid referral growth. That growth is driving compute demand, while proactive assistants could require substantially more capacity and investment.
Preview · 18 more points in the app
Read the Full SummaryKey Takeaways
The host sees personal agents as a potentially bigger software opportunity than code generation was a year earlier, while the guest predicts that the internet and the way most people interact with software will change substantially in the coming years.
The guest describes reservation agents that could check sites every five seconds across restaurants and major cities, and a booking model that could match birthdays or anniversaries with restaurants seeking to prioritize those occasions rather than relying only on first-come, first-served access.
The guest argues that easier AI interfaces could reshape travel agencies, while established businesses may retain value through their data and networks. He also says travel accounts for 50% of platform transaction volume and cites boutique hotels that may pay up to 30% per delivered transaction, as examples rather than a forecast of Instinct’s take rate.
The guest identifies a potential conflict between users’ interests and ad-supported platforms such as Google, TikTok, Instagram, and Snapchat, which may steer purchases through paid advertising. He contrasts that concern with platform monetization examples ranging from Shopify’s estimated 2.5%–3% take rate and Amazon’s rate of up to 10% to Apple’s 30% in-app purchase fee, while saying Instinct’s eventual position is unknown and 30% would be unrealistic for many businesses.
TECHNOLOGY
“It seems like we're in a moment with personal agents that is very similar and arguably probably bigger than what we saw with code generation a year or so ago.”—
“I do think that in the coming years, I think the internet's going to be going to be rewritten…”—
“But then now when you have an agent, you can theoretically just have the agent check every 5 seconds…”—
Instinct and the race for personal AI
Instinct founder describes a personal assistant designed to act through familiar channels such as text, calls, and email, arguing that useful AI should feel simple and socially aware.
What people are using AI agents for
Users apply Instinct to wardrobe planning, shopping, personal finance, subscription cancellation, meeting coordination, and group plans, revealing practical agent use cases beyond one-off prompts.
Rethinking travel, reservations, and the internet
The guest argues agents can reshape reservations and travel by coordinating preferences and logistics end to end, while creating new ways for businesses and customers to match.
Trust, privacy, and personal data
Instinct says users build trust gradually by sharing data at their own pace, while layered controls, content firewalls, and action monitoring aim to protect sensitive information.
The business model behind Instinct
The guest favors a transaction-based model over advertising because agents should act for users rather than steer purchases, and sees distribution as a source of value for merchants.
How existing businesses will adapt
Agents could increase transactions while reducing time spent in apps, so businesses should assess revenue dependence on attention and test agent integrations through controlled experiments.
Designing a personal assistant people love
Instinct prioritizes understandability and a pleasant interaction over showcasing raw capabilities, using staged rollouts, user feedback, and evaluations to refine the experience.
Growth, compute, and competing with Big Tech
Referral growth is accelerating without paid marketing, but Instinct faces difficult compute purchasing decisions; the guest expects proactive assistants to require far more capacity than coding tools.
What’s next for Instinct and personal AI
The guest envisions increasingly simple, adaptive assistants pursuing longer-term goals, handling routine business operations, and interacting through existing channels while maintaining user control and safety.
Related Episodes
Important Disclaimers
Not Investment Advice: For education only — not financial, tax or legal advice.
AI-Generated: Made by AI from source transcripts and may contain errors.
No Guarantees: Past performance does not predict future results; investing carries risk.
Our Commentary · Request Removal