The harder cases

Cognitive augmentation divides by rate.

Now the test runs on the cases where the verdict is live today.

Neural interfaces that restore

Medical neural prosthetics restore what was lost. Paralysis. Sensory loss. Motor disability. They run cleanly through the test.

The interface supplies at a rate your modelling can take in. Your control, applied to the interface, brings the corridor back to the width your build had before the loss.

A paralysed person forms an intent. The interface reads it and moves a prosthetic limb. That intent is the person’s own, working at the rate their modelling has always worked at. The interface widens the corridor. The matching holds.

The verdict is maintenance.

Augmenting the mind

Cognitive augmentations are different. They are interfaces built to give your modelling computing power it would not have produced alone, at rates it has never worked at. They run through the test differently. The test reads the rate.

Some supply below the grain of your modelling. Retrieval. Recall. Information your own record could in principle have supplied alone, only more slowly. These run cleanly. Your control exceeds the drift rate. Your modelling takes it in. The matching holds.

Some supply faster than your modelling can take in without breaking the loop. The matching fails. The loop breaks, at the grain where the augmentation works.

The verdict is parasitic, at the level of the build.

Most transhumanist proposals

Most transhumanist proposals in public debate fail the matching.

Proposals to give your modelling bandwidth orders of magnitude beyond what it has ever worked at. Proposals to join your build to computing systems whose drift rates exceed your control. Proposals to extend your record into external storage, at rates your modelling cannot take in.

These fail the test. Not because they are ambitious. Because they imagine an upgrade the structure does not produce.

Your build is not a subsystem that can be upgraded by adding bandwidth. It is the build your loop runs on. Bandwidth beyond what the loop can take in does not upgrade the loop.

The bandwidth shatters the loop.

That is why the chapter does not endorse the strongest transhumanist proposals.

The objection is not bioconservative. It does not come from guarding the natural body. The chapter has already found organ donation, passive prosthetics and working neural interfaces permissible.

The objection is structural. An augmentation that outruns your control does not produce an upgraded person. It produces a build that has lost the very thing that made it an operator in the first place.

The structure refuses augmentations of that kind, whatever institutional or commercial pressure surrounds them.

Source. Ø Applications, Chapter 8 — Transhumanism and the Augmented Window. The book in full: the420code.org/Applications.pdf. The short edition: the420code.org/models/applications/08-transhumanism-and-the-augmented-window/. The film, Bandwidth does not upgrade you: the420code.org/films/applications/. Its kill switches: APP-8.1 to APP-8.5.