Episode #85: The Conspiracy Theory That Isn't: When Silicon Valley Quietly Changes the Deal

In this episode of Stewart Squared, host Stewart Alsop is joined by his father Stewart Alsop II to cover a wide range of topics sparked by a growing frustration with Anthropic's recent changes to their subscription model, which leads into a broader conversation about trust in Silicon Valley and the historical patterns of companies like Microsoft, Meta, and OpenAI either earning or burning customer loyalty. The two also get into the competitive dynamics between Apple, Google, and Anthropic in the LLM space, LinkedIn's "Browsergate" controversy, the role of IT departments in an AI-driven world, the RISC-V open-source instruction set architecture and its implications for the US-China tech rivalry, the ongoing transformation of the auto industry around EVs and Chinese competition, and whether the growth imperative still holds for the new wave of AI-enabled one- or two-person businesses.

Timestamps

00:00 - Stewart feels suckered by Anthropic's pricing shift, moving from $20 to $200 subscription only to face new usage limits and unexpected charges.
05:00 - Anthropic versus OpenAI trust comparison, with OpenAI buying a podcast signaling lack of strategy while Anthropic remains focused on its original mission.
10:00 - Microsoft's historical distrust traced to MS-DOS licensing deal with IBM, Bill Gates' purely transactional mercantile approach alienating consumers permanently.
15:00 - Apple positioning itself as neutral LLM platform, partnering with Google Gemini embedded at system level while letting users choose their AI.
20:00 - Anthropic compared to early Microsoft serving programmers, while OpenAI risks everything on ego-driven moves despite massive funding rounds.
25:00 - RISC-V open source instruction set architecture origins at Berkeley, China's strategic acquisition of it through Switzerland, semiconductor choke points examined.
30:00 - Microsoft's three CEO eras analyzed, Nadella making IT departments king while Apple cultivated direct consumer trust through Jobs and Cook.
35:00 - Cloud storage and API automation replacing traditional IT gatekeepers, COBOL legacy systems being translated by Claude into modern languages overnight.
40:00 - One-person GLP-1 drug company doing 1.8 billion revenue challenges growth imperative assumptions about venture capital and company scaling.
45:00 - Tesla's lack of model years creating customer engagement problems, Chinese EV dominance threatening legacy automakers still building on gas platforms.
50:00 - Ford rebuilding EV manufacturing from ground up, autonomous vehicles facing real-world infrastructure limitations beyond urban environments.

Key Insights

1. Anthropic has built genuine trust among its users compared to competitors like OpenAI and Meta, but that trust is now being tested. The host feels deceived after being upsold to a $200 monthly subscription, only to find usage limits tightening unexpectedly. This sense of betrayal is significant because trust is the foundation of Anthropic's brand identity and competitive advantage.
2. Trust is the single most important strategic asset a technology company can hold. Companies like Microsoft and OpenAI have historically undermined user trust through mercantile or erratic behavior, while Apple consciously built trust into its culture under Tim Cook, turning it into a durable business advantage that competitors have struggled to replicate.
3. Microsoft has never genuinely earned consumer trust, dating back to its early DOS licensing moves. Its core customer has always been the enterprise IT department, not the end user, which is why consumer-facing products like its digital wallet failed and why users have long resented being subordinated to IT gatekeepers who prioritize control over usability.
4. Apple's emerging strategy positions it as a neutral, trusted platform layer for AI, potentially allowing users to choose their own large language model the way they choose a browser. By partnering with Google on Gemini at the system level while remaining open to other providers, Apple avoids the capital cost of training its own foundation models while leveraging its deep consumer trust.
5. Anthropic's greatest contribution may be enabling ordinary people to write software without technical backgrounds. By focusing on programmers first and then making programming accessible to non-programmers, Anthropic shifted the entire conversation around who can build technology and effectively democratized software development.
6. Legacy enterprise IT departments face an existential threat from AI. The traditional bottleneck of having IT mediate between business needs and technical implementation is dissolving as non-technical employees can now build their own applications. Companies that fail to adapt their internal structures around this reality risk falling behind competitors who embrace AI-driven agility.
7. The electric vehicle industry mirrors the broader technology landscape in that companies built from the ground up around a new paradigm outperform those retrofitting old infrastructure. China and companies like Rivian, which designed EVs without legacy constraints, have structural advantages over traditional automakers who tried to electrify existing gas-car platforms.