Supply and Demand Is Not What Most People Think

4 hours ago

The Supply and Demand model is about the relationship between supply and demand curves, not just quantities or prices.
Changes in quantity alone do not predict price changes; the cause—whether a shift in demand or supply—determines the outcome.
Similarly, changes in price alone do not predict quantity changes; the reason behind the price shift matters.
Many experts, including economists and LLMs like ChatGPT, misunderstand this model, leading to incorrect predictions.
The model predicts equilibrium changes from shifts in curves, not simple cause-effect between price and quantity.

Token prices are currently in a state of supply crunch and instability, with all variables in play, leading to an uncertain future equilibrium.
Supply is increasing due to massive data center and semiconductor investments, improved inference efficiency, and variable token efficiency in new models.
Demand surge is driven mainly by software development, a relatively small field, leaving future use cases, scale, and token needs unknown.

Tautologies, often dismissed as simplistic definitions, are crucial in economics for promoting clear thinking and revealing plausible causal relationships.
The Equation of Exchange (M*V=P*Y) is less intuitive than the Cambridge Equation (M=k*P*Y), which clarifies that NGDP is determined by base money and the share of income held as cash balances.
The savings equals investment (S=I) identity shows that increased saving leads to increased investment, contradicting fears that thrift causes depressions; Keynes' paradox of thrift relates to hoarding, not saving.

Futarchy uses prediction markets to guide decisions, but it fails when markets reflect correlation rather than causation.
A key flaw emerges when one market is canceled after the other resolves, distorting incentives and prices.
In an experimental market with two coins (A with 60% heads, B with uncertain bias averaging 59%), coin B's price rose to $0.90 despite a lower true probability, confirming the bias.

Performance improvements in benchmarks may not reflect live production due to noise in metrics.
Multiple statistics (mean, median, p95, p99) can tell contradictory stories about the same data change.
Cumulative distribution functions (CDFs) visualize the entire distribution, revealing shifts and crossing points.

World models, unlike large language models, learn dynamics from observation and simulate outcomes to predict behavior in the physical world.
Key figures like Yann LeCun, Demis Hassabis, Sam Altman, Fei-Fei Li, and Jensen Huang are advancing world models through different approaches, though definitions vary.
Current AI applications in Earth science (e.g., fire detection, weather forecasting, flood prediction) have improved but struggle with complex systems like hurricanes and the carbon cycle.

Token prices are currently in a state of supply crunch and instability, with all variables in play, leading to an uncertain future equilibrium.
Supply is increasing due to massive data center and semiconductor investments, improved inference efficiency, and variable token efficiency in new models.
Demand surge is driven mainly by software development, a relatively small field, leaving future use cases, scale, and token needs unknown.

Tautologies, often dismissed as simplistic definitions, are crucial in economics for promoting clear thinking and revealing plausible causal relationships.
The Equation of Exchange (M*V=P*Y) is less intuitive than the Cambridge Equation (M=k*P*Y), which clarifies that NGDP is determined by base money and the share of income held as cash balances.
The savings equals investment (S=I) identity shows that increased saving leads to increased investment, contradicting fears that thrift causes depressions; Keynes' paradox of thrift relates to hoarding, not saving.

Futarchy uses prediction markets to guide decisions, but it fails when markets reflect correlation rather than causation.
A key flaw emerges when one market is canceled after the other resolves, distorting incentives and prices.
In an experimental market with two coins (A with 60% heads, B with uncertain bias averaging 59%), coin B's price rose to $0.90 despite a lower true probability, confirming the bias.

Performance improvements in benchmarks may not reflect live production due to noise in metrics.
Multiple statistics (mean, median, p95, p99) can tell contradictory stories about the same data change.
Cumulative distribution functions (CDFs) visualize the entire distribution, revealing shifts and crossing points.

World models, unlike large language models, learn dynamics from observation and simulate outcomes to predict behavior in the physical world.
Key figures like Yann LeCun, Demis Hassabis, Sam Altman, Fei-Fei Li, and Jensen Huang are advancing world models through different approaches, though definitions vary.
Current AI applications in Earth science (e.g., fire detection, weather forecasting, flood prediction) have improved but struggle with complex systems like hurricanes and the carbon cycle.

Hasty Briefsbeta