Books & Sources
23 sources in the library. Add what you are reading, with the passage that made you stop.
Charts, figures and historical documents cited across the Agora, each with its provenance.
Challenges Above a threshold, further growth in output no longer translates into reported wellbeing.
Provenance Maddison Project Database & World Happiness Report · 2023
Method Real GDP per capita indexed to 1972 (2017 PPP); life satisfaction is the global population-weighted mean on a 0–10 Cantril ladder. Series are not causally identified.
Added by @amara
Supports Tutoring quality tracks model capability, which is currently a predictable function of compute — so personalisation improves as systems scale.
Provenance Replication of published scaling-law figures, three model sizes · 2024
Method Median loss across held-out evaluation tasks, log-log axes, power-law fits of the form L = AC^-α. Loss is a proxy for capability, not for pedagogy.
Added by @kenji
Challenges Geography sets the stage but not the outcome — identical terrain diverges under different institutions.
Provenance Geographic commission survey, archival scan · 1887
Method Hand survey at roughly 1:50,000. Useful as a control: one basin, one climate, one population, two sets of institutions.
Added by @noor
Timestamped passages of video and audio, quoted for the exact argument they make.
Supports Rich economies can raise wellbeing while holding material throughput flat.
“The question is not whether we can grow, but whether the things we are growing are the things we said we wanted.”
…if you hold throughput constant and keep improving allocation, you still get better outcomes for two or three decades. That is not degrowth as austerity; it is growth measured in something other than tonnage.
Provenance Public university lecture series, recorded · 2022
Added by @amara
Challenges Adaptive systems standardise instruction because they optimise the measurable part of learning.
“A tutor's central act is deciding what not to teach today. No system optimising for completion will ever make that decision.”
…the human tutor withholds. They see that the student is tired, or defensive, or one conversation away from giving up, and they slow down. Completion metrics punish exactly that behaviour.
Provenance Learning sciences seminar, recorded · 2023
Added by @kenji
Book
Jared Diamond · 1997
A geographic account of why some societies industrialised earlier, centred on crops, domesticable animals and continental axes.
Book
Daron Acemoglu & James A. Robinson · 2012
Places inclusive versus extractive political institutions, rather than geography or culture, at the centre of long-run prosperity.
Book
Nicholas Carr · 2010
Argues that networked media reshapes reading into scanning, trading depth of comprehension for breadth of exposure.
Long arguments, read slowly and together.
Book
Amartya Sen · 1999
Argues that development should be measured as the expansion of real freedoms people enjoy, not as income growth alone.
Book
Jared Diamond · 1997
A geographic account of why some societies industrialised earlier, centred on crops, domesticable animals and continental axes.
Book
Daron Acemoglu & James A. Robinson · 2012
Places inclusive versus extractive political institutions, rather than geography or culture, at the centre of long-run prosperity.
Book
Daniel Kahneman · 2011
Two systems of thought — fast, intuitive judgement and slow, deliberate reasoning — and the predictable errors of the first.
Book
Hannah Arendt · 1958
Distinguishes labour, work and action, and defends the public realm as the space where human plurality becomes visible.
Book
Nicholas Carr · 2010
Argues that networked media reshapes reading into scanning, trading depth of comprehension for breadth of exposure.
Book
Adam Smith · 1776
The founding argument that specialisation and exchange, not hoarded bullion, are the sources of a nation's wealth — read here against its own moral philosophy.
Book
James C. Scott · 1998
How schemes to improve the human condition fail when they replace local, practical knowledge with legible administrative abstractions.
Book
Donella H. Meadows · 2008
A short, generous introduction to stocks, flows, feedback and leverage points — the grammar behind most arguments about growth and limits.
Book
Thomas S. Kuhn · 1962
The account of paradigms and anomalies that reframed science as a social practice punctuated by crisis, not a smooth accumulation of facts.
Book
Abhijit V. Banerjee & Esther Duflo · 2011
Field experiments on how poor households actually decide about food, schooling, health and credit — an argument for small, testable claims over grand theories.
Book
David Graeber · 2011
An anthropological history arguing that credit preceded coinage, and that debt is a moral language before it is an economic one.
Book
Michel Foucault · 1975
The birth of the prison as a model for modern institutions: power exercised through visibility, measurement and the ordinary timetable.
Book
David Deutsch · 2011
A defence of good explanations as the engine of unlimited progress, and of optimism as an epistemological position rather than a mood.
Evidence as it was actually reported.
Research paper
Benjamin S. Bloom · 1984
Reports that tutored students outperformed conventionally taught peers by roughly two standard deviations, framing the challenge of scaling tutoring.
Research paper
Richard A. Easterlin · 1974
The original statement of the Easterlin paradox: within countries richer people report more happiness, yet national happiness does not rise with long-run growth.
Research paper
L. Chen, K. Tokoro · 2024
A meta-analysis of 62 studies on interruption cost, arguing attention behaves less like a stock and more like a recovering tissue.
Research paper
Vaswani et al. · 2017
The transformer paper: the architecture behind current language models, cited here because most arguments about AI and knowledge rest on it.
Reporting and essays worth citing.
Article
Rosalind Grant · 2025
Field reporting from three school systems piloting adaptive learning software, and the gap between the demo and the classroom.
Article
Richard S. Sutton · 2019
Seven hundred words arguing that general methods which scale with computation beat hand-built human knowledge — the most quoted essay in the AI circle.
Arguments made aloud.
Podcast / Lecture
Dr. Femi Akintola & Amara Moyo · 2026
A 74-minute exchange on whether infrastructure or institutional reform should come first in landlocked economies.
Podcast / Lecture
Noémie Simard · 2025
Fifty minutes on why disagreement is not a failure of the public realm but its precondition.
Podcast / Lecture
David Foster Wallace · 2005
Twenty minutes on attention as a moral discipline: what you choose to notice is what your adult life is made of.