In the current debate between the Trump II Administration and the US Federal Reserve, maybe we are not doing enough thinking outside the box. The Trump II Administration has formed an unusual alliance with Big Tech believing, in general, that all government functions could be replaced by a cell phone app maybe connected to a Cloud Computer--and the Federal Reserve performs a government function. Thinking out of the Box about the Federal Reserve brings up a number of deeper questions that are explored in this posting.
My first step was to ask ChatGPT whether the Fed could be replaced by a computer:
ChatGPT was willing to go further and map out the process for a "Fully Algorithmic Federal Reserve" and called it FedOS 1.0!
The graphic above is similar to the one produced by ChatGPT but I've added VOTING and made the process more consistent with Systems Theory. Here are the details:
- MANDATE Laws passed by Congress (the Banking Act of 1935 and the Dodd-Frank Reform 2026) defined a REACTION FUNCTION that must be minimized by the Fed --essentially the Taylor Rule. Thus, current Fed Policy is Rule Based.
- DATA The Fed currently gathers all the data needed for decision making: Inflation, Real Activity, Financial Stability and Global Impacts.
- VALIDATION There is a problem with collecting data in real time: the data is noisy and has to be validated.
- MULTI MODEL There is no one perfect model of the economy (what ChatGPT calls the "Big Brain"). The multi-model, ensemble approach to forecasting is also used by Hurricane forecasting (see the Boiler Plate). I would recommend using the Akaike Information Criterion (AIC) to evaluate models.
- REGIME CLASS There are different regimes that the FED operates under in addition to Business-as-Usual (BAU). In another post (here) I statistically identify those regimes as BAU, Unemployment Shock, Banking Crisis and Inflation Shock.
- POLICY ENGINE The Policy Engine would present multiple models predicting FED actions under different regimes.
- REACTION FUNCTION would describe the best action under different regimes (here). It describes a desired state (Q*) to be compared with an actual state (Q). (Q*-Q) is a Cybernetic Reaction Function used to control systems.
- VOTING Because I think Big Tech basically believes in Democracy (maybe, at least many of the employees) and is making a mistake supporting anti-Democratic politicians, I have allowed for voting using a cell phone app to provide democratic input to the FOMC (similar to the Atlanta Fed EconomyNow App, see below in the Notes, but with voting input).
- UNCERTAINTY The main purpose of the FOMC is to allow human input, accountability and to avoid technocratic mistakes under uncertainty.
- OVERSIGHT Oversight involves identifying regimes that are not BAU.
- OVERRIDE Under conditions of uncertainty, the FOMC can override the Policy Engine.
- POLICY Ultimately, the FOMC decides Fed policy using all available information.
- COMMUNICATE The FOMC is charged with transparently communicating the policy decision and explaining their evaluation of existing data and output from the Policy Engine. The communication would include open evaluation of how well the Fed has performed under different policy regimes.
So, ultimately and obviously, the FOMC cannot and should not be replaced by a computer. However, a process can be set up that uses all available multi-model computer input and public voting to allow the FOMC to make transparent policy decisions.
ChatGPT reports that:
I'm not so sure. US Executive Agencies (the Department of Defense, HHS, Homeland Security, Dept. of State, Immigration, the White House, Congress, the Supreme Court, etc., etc.) have all been compromised by poor leadership under an authoritarian political regime (the Trump II Administration). Democratic voting itself has been under threat and is too slow and indirect a process to prevent abuse. Indeed, the entire system seems quite Anachronistic.
The FedOS 1.0 model provides a general approach to reforming failing institutions. The FOMC as it currently functions is quite close to FedOS 1.0 and could be used as a model. If the US Supreme Court blocks the Trump II Administration from destroying the Fed, the model would be easy to implement by a progressive Political Regime. I would encourage Big Tech to start supporting "institutional reform" rather than supporting incompetent Right-Wing Administrations and politicians in the hope that these administrations will be "good for business".



No comments:
Post a Comment