Friday Assorted Links: Tokenized Compute, Dulles Airport Design, Amazonian Earthworks, and More
Key Takeaways
- •Bob Gelfond's Substack post explores tokenizing economic value from computational resources as AI-driven GPU demand intensifies interest in liquid compute markets.
- •A White House post on X has raised concerns about potential alterations to the Eero Saarinen-designed main terminal at Washington Dulles International Airport, a mid-century modern architectural landmark opened in 1962.
- •A Nature study documented more than 20,000 precolonial earthworks in southwestern Amazonia, providing new evidence of large-scale, organized Indigenous settlements before European contact.
- •An analysis shared on X examines what AI company valuations can be justified on fundamental grounds as venture capital funding reaches record levels.
- •Economist Matt Clancy draws on Industrial Revolution-era lessons to identify structural barriers that make sustaining extremely rapid economic growth persistently difficult.
Writing at Marginal Revolution, economist Tyler Cowen presents a weekly collection of noteworthy links spanning technology, urban policy, economics, and archaeology.
1. Tokenizing the value from compute?
A Substack post by Bob Gelfond explores the concept of tokenizing the economic value derived from computational resources, a question that intersects with ongoing efforts to create liquid markets for compute capacity as AI training workloads drive surging demand for GPUs and cloud infrastructure.
2. Will they ruin the look of Dulles airport?
A post from the White House on X raises questions about potential changes to the iconic design of Washington Dulles International Airport, whose main terminal was designed by the renowned Finnish-American architect Eero Saarinen and opened in 1962. The Saarinen terminal is widely regarded as a landmark of mid-century modern architecture, notable for its soaring catenary-curved roof and the original mobile lounge system designed to transport passengers directly to aircraft.
3. Pathways into homelessness.
Economist Paul Novosad shares research findings on X examining the various routes through which people fall into homelessness.
4. Defensible valuations for AI companies.
An analysis shared on X considers what valuation levels for artificial intelligence companies can be justified on fundamental grounds amid the sector's rapid expansion, a question that has grown more pressing as venture capital funding for AI startups has reached record levels.
5. Further reasons why very rapid economic growth is very difficult to achieve.
Economist Matt Clancy examines on his blog additional structural barriers that make sustaining extremely high rates of economic growth a persistent challenge, drawing on historical lessons from the Industrial Revolution era.
6. Over 20,000 precolonial earthworks in the Southwest Amazonia.
A study published in Nature documents the discovery of more than 20,000 precolonial earthworks in southwestern Amazonia, providing new evidence of the scale and complexity of Indigenous settlements in the region before European contact. The findings add to a growing body of research, aided by lidar and remote-sensing technologies, that has been progressively reshaping archaeological understanding of dense, highly organized pre-Columbian societies throughout the Amazon basin.