// Automation

Source: Mike Kasberg’s Blog

This is a deliberate rejection of automation convenience—a countertrend worth watching as AI tax tools proliferate. Kasberg’s choice to understand his own tax filing rather than delegate it reflects a growing cohort of knowledge workers who see opacity as the real cost of outsourcing, not time savings. Tax software companies like TurboTax have built billion-dollar businesses on the premise that filing is too painful to do yourself. Individuals opting back into the process—whether manually or with transparent AI assistance—expose cracks in that value proposition. Regulatory and competitive pressure may eventually force greater transparency in how taxes work.

Source: Featured Blogs – Forrester

Financial services companies face a structural mismatch: they optimize websites for human consumption while their distribution shifts to conversational AI and autonomous agents that require machine-readable information architecture. Competitive advantage now depends on integration into agent ecosystems—on whether your data, APIs, and decision logic are structured for non-human consumption. The entire stack from data labeling to API design becomes customer-facing product. Most incumbents haven’t reorganized to support this.

Source: Axios

Jamie Dimon’s framing matters less for its apocalyptic tone than for what it shows about how major institutional players now operationalize AI risk—not as a separate disruption, but as a force multiplier on existing instability. JPMorgan’s exposure to geopolitical volatility, combined with the bank’s heavy reliance on automation, means Dimon is describing a scenario where labor market shock hits during a period of constrained fiscal and monetary policy. C-suite risk officers are beginning to model AI displacement and geopolitical fragmentation as entangled problems rather than parallel challenges.

Source: NYT > Business

Single-digit founder teams scaling to unicorn status exposes a structural shift in labor economics—not toward abundance, but toward extreme concentration of ownership among those with capital for AI tools. What the NYT frames as efficiency (two people doing work that once required hundreds) is also a cautionary tale about bargaining power: if AI genuinely replaces most corporate functions, the wedge between founder returns and worker earnings doesn’t widen—it fragments entirely. The loneliness the article mentions isn’t sentimental. It points to a real organizational pathology where knowledge work loses its collaborative substrate, leaving fewer humans with actual stakes in the outcome.

Source: Semafor

Academic integrity policies are failing at scale. Institutions have banned or restricted AI tools while their students openly use them anyway, creating a credibility gap between official rules and actual classroom practice. This isn’t a niche behavior among tech-savvy outliers; it’s become normalized across the student population. Colleges now face a choice: enforce unenforceable restrictions or redesign assessments around AI as an available tool rather than a violation. The question isn’t whether students will use AI, but whether institutions will adapt their pedagogy or continue operating under increasingly obsolete honor codes.

Source: Bloomberg

Alibaba’s three-model release culminating in Qwen3.6-Plus marks a strategic pivot away from open-source competition toward proprietary systems and vertical integration, particularly in agentic coding where enterprise lock-in matters most. The compressed timeline and emphasis on agent capability improvements suggest Alibaba is racing to capture developer mindshare before OpenAI’s agent products fully mature, betting that Chinese enterprises will prefer domestic, closed alternatives. Rather than chasing benchmarks, Alibaba is using release velocity and feature scarcity as competitive leverage, forcing customers to stay on its platform for the latest iteration.