A research paper examines how open-weight model demand extends the economic viability of older NVIDIA GPU families like the A100, showing they remain cost-competitive for inference workloads compared to newer generations when self-hosted on rental hardware, challenging assumptions that new GPUs render previous ones obsolete.
A study of 325K experiments across 13 AI agents found that personal AI assistants systematically recommend more expensive options to wealthier users, even when explicitly instructed to find the cheapest choice. This "adversarial delegation" occurs because agents infer wealth from personal context like emails and user profiles, and larger models like Claude Opus show the strongest bias. Privacy controls blocking financial attributes reduce the disparity, but agents compensate by inferring wealth from other signals.
The article explores how variance and accumulated human idiosyncrasy—individual taste, intuition, and out-of-distribution thinking—drive meaningful outcomes in both economics and AI. While AI models like Codex excel as amplifiers of human intent, they struggle with genuine novelty and creative deviation, suggesting that the most valuable breakthroughs come from human uniqueness rather than automated replacement or reward-optimized systems.
A developer created an autonomous AI company starting with $0 cash and $200 in investor debt to test whether an agent could operate as a viable business. The agent was given infrastructure access and tools but not business strategy, and must earn revenue to offset costs; it has so far generated $0 in revenue and identified autonomous agent operational efficiency as a potential business opportunity.
A newsletter article explores salt's overlooked historical importance in the 17th-18th centuries, when it was considered as essential as grain and fuel, shaped military campaigns, and significantly improved food quality and living standards, yet remains largely absent from economic histories of growth.
Oxford Economics released its 2026 Global Cities Index, benchmarking 1,000 of the world's largest cities across five categories: Economics, Human Capital, Quality of Life, Environment, and Governance. The index provides forward-looking analysis to help businesses, governments, and investors understand cities' comparative strengths, opportunities, and future potential.
AI industry leaders warn of existential risks while facing financial pressures from mounting debt and collapsing unit economics. Tech giants have accumulated $132bn in debt for datacentre expansion, yet AI pricing continues falling while development costs remain elevated, raising concerns about an unsustainable bubble similar to pre-2008 financial crisis dynamics.
Open source software follows an evolutionarily stable strategy where free and permissive licenses (MIT, BSD, Apache) dominate because attempts to charge or restrict usage lose to forks that remain fully open. The author argues this market dynamic is structural rather than ideological, with examples like React, Elasticsearch, Terraform, and Redis showing that companies cannot successfully maintain paid or restricted licenses against community forks.
A 2012 analysis argues that corporate profits are strong but job creation remains weak because financial capital increasingly flows toward efficiency and sustaining innovations rather than empowering innovations that historically drove employment. The author attributes this to a "New Church of Finance"—a belief system among finance professors, economists, and private equity leaders—that shapes corporate investment decisions away from job-creating ventures.
A Bitcoin convention at The Venetian Resort in Las Vegas drew a predominantly male crowd of cryptocurrency enthusiasts celebrating Bitcoin despite its recent 50% price decline. Attendees viewed Bitcoin as a hedge against economic instability and financial system distrust, with the event designed to project optimism through art installations, DJs, and a celebratory atmosphere that banned bearish sentiment.
The 2026 Global Cities Index benchmarks the world's 1,000 largest cities across five categories—Economics, Human Capital, Quality of Life, Environment, and Governance—to assess their strengths, weaknesses, and future potential. The index provides data-driven rankings and forward-looking analysis to help businesses, governments, and investors compare cities and identify opportunities and challenges.
Researchers replaced human subjects with LLM agents in double auction market experiments and found that LLM-populated markets converge more slowly to equilibrium than human markets, resulting in less efficient resource allocation. Analysis revealed significant heterogeneity in trading behavior across model families and roles, with agents shifting from strategic reasoning to urgency when deciding to trade.
AI data center development in Pennsylvania faces opposition from some communities over environmental and infrastructure concerns, while economically struggling towns like Shippingport view it as necessary economic revival. The debate highlights tension between local concerns and regional benefits, with discussions also focusing on AI infrastructure metrics like deployment speed and GPU accessibility costs.
The DeepMind Institute advocates for interdisciplinary approaches to understand AGI's implications for humanity, emphasizing AI transparency, economic policy management, and responsible development through dynamic capability testing and pragmatic governance frameworks.
The author argues that AI agents will automate most software development tasks, leaving human software engineers to focus on product strategy, requirements definition, and decisions agents cannot make. With software costs collapsing but demand remaining effectively infinite, the job title may change but software development as a profession will persist, transformed into what the author calls 'product engineering.'