• Consider HubSpot (HUBS) as an AI-enabled SMB software company to monitor: its CRM data, workflow tools, and new YouSpot product could support growth, but YouSpot was only about 30 days old and its traction remains unproven.
• For AI software investments, favor companies solving a specific industry or customer workflow and with strong customer reach and retention; avoid relying on a temporary gap in a frontier model’s capabilities.
• The discussion provides no price targets or explicit buy recommendations, so treat these as themes to research rather than immediate trade signals.

AI-Native Software and SMB/Solopreneur Software

• Shah and Schmidt are broadly bullish on the opportunity to build software with AI, arguing that AI makes it faster and cheaper to create, test, and improve products.
• They see room for startups that solve specific industry or customer problems rather than competing directly with frontier AI labs on broad, horizontal applications.
• Shah argues that founders should start with a narrow customer group and problem, get a product in front of users quickly, and learn from actual usage rather than relying on elaborate market forecasts.
• Distribution may be harder as software becomes easier to build, because more products will compete for customer attention.
• The discussion points to a growing market for software serving one-person businesses: Shah cited record U.S. company formations in 2025 and said more than 70% of new businesses were expected to remain one-person companies for at least two years.

Takeaways

• Look for AI software opportunities built around a specific customer workflow, specialized data, or industry context—not just a feature a large AI lab could readily add.
• For early-stage companies, assess how they will reach customers and retain them, not only how quickly they can build a product.
• The transcript presents solopreneur-focused software as a potential growth area, but it does not identify a specific publicly traded investment in that market.

HubSpot (HUBS)

• HubSpot’s original focus was marketing software for small and medium-sized businesses. The company later expanded into CRM after finding that many of its customers were not using one, partly because existing options felt intimidating or expensive.
• Shah says HubSpot’s broader product approach was intended to reduce the fragmentation customers faced across marketing tools.
• The company is adapting its product and technical foundations for AI, including ways to work with both structured data and unstructured customer context.
• HubSpot has launched YouSpot, a CRM aimed at one-person businesses, and Shah said he began testing monetization early, including a consumption-based model.
• Shah argues that AI interfaces such as ChatGPT may change how people access business software, but that the underlying work, customer data, and ability to deliver outcomes are more durable sources of value than the interface alone.
• He also described HubSpot connectors for ChatGPT and Anthropic products, suggesting the company is positioning its software to work through external AI interfaces as well as its own.

Takeaways

• The discussion supports monitoring HubSpot’s ability to turn its customer relationships, CRM data, and broader platform into useful AI-driven workflows.
• YouSpot provides an early test of demand for software tailored to solopreneurs and of consumption-based pricing; Shah said the product was about 30 days old at the time of recording, so the transcript offers little evidence about its longer-term performance.
• The potential for AI interfaces to become a new way customers use business software is both a strategic opportunity and a competitive consideration for HubSpot. Shah’s view is that the interface is not the whole product, but the company still needs to make its data and services valuable through those new interfaces.
• No stock price target or specific recommendation was given.

OpenAI (Private)

• OpenAI is discussed as one of the frontier AI labs moving into applications and building products that could become a primary interface for users.
• Shah argues that founders should not assume a gap in a current OpenAI product will remain open; in his view, “it can’t do that yet” is not a durable business thesis.
• He says OpenAI and similar companies may sell AI capabilities directly to consumers and large enterprises, while also creating opportunities for software companies that package AI for particular industries or workflows.
• He considers it less likely that a frontier lab would directly pursue every specialized software market, such as CRM, because of competing priorities and incentives.

Takeaways

• The transcript presents OpenAI as a powerful infrastructure and distribution force, but not as the only potential beneficiary of AI adoption.
• For private-market investors, the discussion suggests weighing the labs’ ability to expand into applications against the opportunity for focused software companies to use their models.
• OpenAI is private; the transcript gives no valuation, price target, or investment recommendation.

Anthropic (Private)

• Anthropic is mentioned alongside OpenAI as a frontier AI company expanding into applications and as a provider of models that businesses can connect to.
• HubSpot has built a connector for Anthropic products, reflecting the strategy of making business software accessible through external AI interfaces.
• Shah argues that model choice is becoming more flexible: companies can route different tasks to different models and switch providers, reducing dependence on any single supplier.

Takeaways

• The discussion favors a flexible AI ecosystem in which application companies can use multiple models rather than rely on one provider.
• For investors considering private AI companies, the transcript does not make a direct case for Anthropic’s valuation or prospects; it focuses more on the opportunity for businesses built on top of model providers.
• Anthropic is private, and no price target or specific recommendation was given.

Salesforce (CRM)

• Shah said that, in a survey of HubSpot customers conducted several years before the recording, 20% used Salesforce, 20% used another CRM, and 60% did not use a CRM.
• He described CRM as an adjacent market for HubSpot and noted that established CRM vendors could move further into marketing software.

Takeaways

• The discussion highlights both the potential to attract businesses that have not adopted CRM and the competitive overlap between CRM and marketing software.
• Investors may want to consider how established CRM providers respond as AI changes product interfaces and workflows. The transcript does not offer a specific view on Salesforce’s stock or a price target.

Amazon / AWS (AMZN)

• AWS and cloud computing are used as an analogy for AI: Shah compares AI models to infrastructure that other companies can build on, much as software companies built products using cloud services.
• He argues that a wider choice of commercial and open-source models makes it less likely that one provider will control the essential input for every AI software company.

Takeaways

• The analogy suggests that infrastructure providers may benefit from broad adoption, while application companies can also create value on top of the infrastructure.
• AWS is an Amazon business, but the transcript does not discuss Amazon’s stock, financial outlook, or a specific investment recommendation.

Intuit (INTU)

• Intuit is cited as an example of a company that built a large business serving SMBs. Shah used it to counter the historical skepticism that small-business software could support a multibillion-dollar company.

Takeaways

• The discussion supports the view that SMB software can be a substantial market, particularly when products solve practical business needs and reach a large customer base.
• Intuit is mentioned as an example, not as a specific stock recommendation; no price target or company outlook is provided.

DoorDash (DASH) and Uber (UBER)

• DoorDash and Uber Eats are used to illustrate that a user-facing app may not be the main source of a business’s value.
• Shah points instead to the underlying network, restaurant relationships, delivery logistics, and ability to complete the service as the harder-to-replicate elements.

Takeaways

• The discussion suggests examining the operational assets and networks behind a digital interface, rather than judging a company only by its app.
• This is an illustrative point about business durability, not a specific bullish or bearish view on DoorDash or Uber stock.

Twitter / X (Private)

• Twitter’s early developer ecosystem is described as open and productive, with many third-party products built using its APIs.
• Shah says the platform later became less friendly to developers, illustrating how platform owners can damage trust by restricting or changing access.

Takeaways

• For platform businesses, developer trust and predictable access can help attract complementary products; reducing access can weaken that ecosystem.
• X is not a separately publicly traded stock. The transcript uses Twitter as a historical example and gives no investment recommendation.

Workday (WDAY)

• Workday is mentioned as an example in a prior essay about the challenge established software companies face when adapting to AI-native products.
• The transcript does not provide further analysis of Workday’s business or technology strategy.

Takeaways

• The reference raises the general question of whether incumbents can adapt their software and processes to AI, but it is not a specific assessment of Workday’s investment outlook.
• No price target or recommendation was given.

AI Software Infrastructure and Developer Ecosystems

• The discussion presents AI models as increasingly infrastructure-like, with multiple commercial and open-source options and tools that can route tasks among models.
• Shah says business software may retain value even if users increasingly interact through AI assistants, because the software can hold customer data, manage workflows, and deliver outcomes.
• He also emphasizes that platform companies can build durable ecosystems by making it easier for developers to integrate, launch products, and earn revenue.

Takeaways

• Potential investment areas discussed include AI-enabled business software, workflow automation, specialized data and context, and tools that help developers build on AI platforms.
• The transcript’s central caution is that a product gap alone may not be defensible: assess whether a company has a durable customer problem, useful data or workflow position, and a credible way to reach customers.