• Instinct is private, so it is not a directly investable public stock; treat its reported growth and transaction figures as unverified, and watch for evidence of durable revenue and manageable compute and security costs.
• The AI-agent theme remains speculative: monitor whether agents reliably generate transactions and whether public companies can benefit without losing customer relationships or economics.
• For Alphabet (GOOGL/GOOG), Meta (META), and Snap (SNAP), track changes in user engagement and ad exposure, but the discussion provides no basis for a specific trade.
• Monitor Amazon (AMZN), DoorDash (DASH), Uber (UBER), and Lyft (LYFT) for agent-driven ordering and booking adoption; easier access could boost transactions even as their apps become less central.

Instinct (Private company)

• Noah Shinn described Instinct as a personal AI agent that acts through familiar channels such as text, email, and calls, rather than requiring users to adopt a new app.
• Shinn reported over $1 billion in annual transaction volume on a small, invite-only user base, with about 50% coming from travel. He also cited 10%–11% daily growth, driven by word of mouth, and said the company had spent $0 on marketing.
• The company is exploring a transaction-based business model: users could receive the service for free while merchants pay for distribution. Shinn said the potential take rate is uncertain and depends on how much value and distribution the agent provides.
• Shinn referred to a latest round of roughly $1 billion at an approximately $10 billion valuation, and named Sequoia, Benchmark, and KOTU as investors.
• Risks discussed include rapidly growing compute needs, the cost and lead time of adding capacity, privacy and security concerns, and reliance on existing communication channels such as iMessage and WhatsApp. Shinn also acknowledged that early product versions lacked some of the safety systems later added.

Takeaways

• Instinct is a private-company opportunity, not a publicly traded stock. The reported growth and transaction figures are company claims from the interview, not independently verified results.
• For investors assessing the AI-agent theme, watch whether user growth converts into durable usage and revenue—and whether compute and security costs allow the service to scale economically.
• The business model is still being tested. The proposed merchant-funded approach depends on agents becoming a meaningful source of transactions and distribution.

AI agents and software interfaces (Investment theme)

• Shinn argued that agents could replace some app-based browsing and form a simpler interface for booking travel, arranging reservations, shopping, and coordinating schedules.
• He described a possible shift away from products that monetize users’ time and attention toward businesses that benefit from more transactions, even if customers spend less time in an app.
• He also suggested that agents could help businesses by reducing friction and increasing purchases or service usage. He expects businesses to test agent integrations gradually rather than immediately changing their entire model.
• No specific public AI-agent stock, compute supplier, price target, or investment recommendation was identified.

Takeaways

• Consider the theme as a potential source of both disruption and growth: app-based interfaces may lose value, while underlying services could gain from easier access and higher transaction volume.
• When evaluating companies, distinguish revenue tied to attention or advertising from revenue tied to delivering the underlying product or service.
• Track whether agents can operate reliably and securely at a cost that supports sustainable economics. The transcript emphasizes that proactive agents may require substantially more compute than simpler AI tools.

Online platforms and advertising businesses

• Shinn cited Google (Alphabet), TikTok (ByteDance), Instagram (Meta Platforms), and Snapchat (Snap) as examples of platforms where free access is supported by advertising and efforts to influence purchasing behavior.
• He argued that agents aligned with users’ preferences could reduce exposure to ads and time spent on platforms, potentially challenging businesses that depend heavily on user attention.
• He also said that less friction could increase purchases or service use, so reduced app engagement would not necessarily mean lower transaction volume for every business.

Takeaways

• The discussion suggests a potential risk to attention- and ad-dependent business models, but it does not establish that these companies’ overall revenues will decline.
• For Alphabet (GOOGL/GOOG), Meta (META), and Snap (SNAP), monitor whether AI agents change ad exposure, user engagement, or advertisers’ access to customers. TikTok’s parent, ByteDance, is private.
• The transcript offers no company-specific financial forecasts or stock recommendations.

E-commerce, travel, and delivery platforms

• Shinn used Amazon (AMZN) as an example of a platform that monetizes commerce partly through upselling. He suggested agents could change how users discover and buy products.
• He described travel booking as a major potential use case, saying travel accounted for about 50% of Instinct’s transaction volume. Agents could use stored preferences to arrange flights, hotels, transport, and activities.
• He discussed Uber Eats and DoorDash (DASH) as examples of ordering services that could face a new interface, but argued that easier ordering might increase transaction volume.
• Shinn similarly suggested that proactive ride booking could make services such as Uber (UBER) and Lyft (LYFT) more convenient and potentially increase usage. He did not cite financial effects for any of these companies.

Takeaways

• The potential impact is mixed: agents may weaken the importance of a company’s app or website while sending it more transactions.
• For Amazon, DoorDash, Uber, and Lyft, watch how they respond to agent-driven ordering and whether they can preserve customer relationships and economics when another interface mediates the transaction.
• Travel agencies and booking platforms were discussed as a sector that could be reshaped, but no specific travel-company stock was named.

Payments and commerce infrastructure

• Shinn mentioned Apple Pay, American Express (AXP), Visa (V), Mastercard (MA), and Stripe as examples of companies or services in the payments stack.
• He characterized payment processing as a relatively small portion of the overall value created in a transaction and said Instinct was more interested in earning a take rate for directing customers to merchants than replacing payment infrastructure.
• Shopify (SHOP), Stripe, Amazon, and Apple’s in-app commission were discussed as examples of different platform take rates. These were benchmarks in a business-model discussion, not company-specific investment forecasts.

Takeaways

• The transcript suggests agents may create a new distribution layer while continuing to use established payment rails; it does not present payments companies as direct targets for displacement.
• Investors may want to monitor whether agents capture meaningful merchant fees without undermining the economics of existing platforms and payment providers.
• No specific take rate for Instinct was set.

Spotify (SPOT)

• Spotify was mentioned as a possible source of information about users’ music tastes, which an agent could use when coordinating plans or recommending activities.
• The discussion did not address Spotify’s financial outlook or suggest that the agent would replace the service.

Takeaways

• The mention points to a possible role for personal data and user preferences in agent-driven recommendations, but provides no standalone investment thesis for Spotify.