FIML is both a practice and a theory. The practice  is roughly described here and in other posts on this website.

The theory states (also roughly) that successful practice of FIML will:

• Greatly improve communication between participating partners
• Greatly reduce or eliminate mistaken interpretations (neuroses) between partners
• Give partners insights into the dynamic structures of their personalities
• Lead to much greater appreciation of the dynamic linguistic/communicative nature of the personality

These results are achieved because:

• FIML practice is based on real data agreed upon by both partners
• FIML practice stops neurotic responses before they get out of control
• FIML practice allows both partners to understand each other’s neuroses while eliminating them
• FIML practice establishes a shared objective standard between partners
• This standard can be checked, confirmed, changed, or upgraded as often as is needed

FIML practice will also:

• Show partners how their personalities function while alone and together
• Lead to a much greater appreciation of how mistaken interpretations that occur at discreet times can and often do lead to (or reveal) ongoing mistaken interpretations (neuroses)

FIML practice eliminates neuroses because it shows individuals, through real data, that their (neurotic) interpretation(s) of their partner are mistaken. This reduction of neurosis between partners probably will be generalizable to other situations and people, thus resulting a less neurotic individual overall.

Neurosis is defined here to mean a mistaken interpretation or an ongoing mistaken interpretation.

The theory of FIML can be falsified or shown to be wrong by having a reasonably large number of suitable people learn FIML practice, do it and fail to gain the aforementioned results.

FIML practice will not be suitable for everyone. It requires that partners have a strong interest in each other; a strong sense of caring for each other; an interest in language and communication; the ability to see themselves objectively; the ability to view their use of language objectively; fairly good self-control; enough time to do the practice regularly.

Wolfram’s ‘computational irreducibility’ explains FIML perfectly

[In mathematics, a ‘computation’ is the process of performing mathematical operations on one or more inputs to produce a desired output. A problem in analyzing human psychology arises when we understand that human psychology cannot be reduced computationally. The ‘computational irreducibility’ of human psychology does not mean, however, that there is no way to probe it and understand it. In the following essay, I show how FIML practice can greatly enhance our understanding of our own psychologies and, by extension, the psychologies of others.

Rather than rely on tautological data extractions or vague theories about human psychology, FIML focuses on small interpersonal exchanges that can be objectively agreed upon by at least two people. These small exchanges correspond to what Wolfram calls ‘specific little pieces of computational reducibility’. When we repeatedly view our psychologies from the point of view of specific little pieces of computational reducibility, we begin amassing a profoundly telling collection of very good data that shows how we really think, speak, and act.]

FIML is a method of inquiry that deals with the computational irreducibility of humans. It does this by isolating small incidents and asking questions about them. These small incidents are the “little pieces of computational reducibility” that Stephan Wolfram remarks on at 42:22 in this video. Here is the full quote:

One of the necessary consequences of computational irreducibility is within a computationally irreducible system there will always be an infinite number of specific little pieces of computational reducibility that you can find.

42.22 in this video

This is exactly what FIML practice does again and again—it finds “specific little pieces of computational reducibility” and learns all it can about them.

In FIML practice, two humans in real-time, real-world situations agree to isolate and focus on one “specific little piece of computational reducibility” and from that gain a deeper understanding of the whole “computationally irreducible system”, which is them.

When two humans do this hundreds of times, their grasp and appreciation of the “computationally irreducible system” which is them, both together and individually, increases dramatically. This growing grasp and understanding of their shared computationally irreducible system upgrades or replaces most previously learned cognitive categories about their lives, or psychologies, or how they think about themselves or other humans.

By focusing on many small bits of communicative information, FIML partners improve all aspects of their human minds.

I do not believe any computer will ever be able to do FIML. Robots and brain scans may help with it but they will not be able to replace it. In the not too distant future, FIML may be the only profound thing humans will both need to and be able to do on their own without the use of AI. To understand ourselves deeply and enjoy being human, we will have to do FIML. In this sense, FIML may be our most important human answer to the AI civilization growing around us. ABN