The discussion begins with signs that fears of an immediate software collapse may have been overstated. Salesforce’s strong performance suggests enterprise platforms can remain valuable if they control trusted data and integrate AI effectively, even as traditional interfaces become less important. The conversation then turns to Meta, where an aggressive effort to reorganize around AI produced more code but limited productivity gains, alongside reliability, security, morale, and cost concerns. The hosts argue that weak incentives and performative AI adoption can undermine otherwise powerful technology. Meta’s responsibility toward young users is examined next, with proposed changes such as removing visible engagement metrics and offering chronological feeds. Yet meaningful reform may conflict with recommendation-driven business models that keep children engaged. Finally, South Korea’s volatile AI-linked stock market illustrates how leverage, limited economic opportunity, and herd behavior can magnify losses. The episode closes by comparing today’s enthusiasm for AI, software, and memory chips with earlier speculative cycles, warning that technological progress can coexist with financial instability and painful disruption.