Learning from Experience

Alpha Function and Artificial Intelligence

It is obvious to say that – and even rather banal to do so at this stage in the game – that we are in danger of losing our capacity to imagine, think, and the motivation to create things for ourselves by ourselves. In this short essay, I hope to do something at least slightly more interesting than that, by providing examples of ways which a few of the psychoanalytic concepts created by Wilfred Bion can be re-vivified in this context in order, not only to offer important perspectives in terms of the perils of handing over various forms of mentation to AI, but also as an essential reminder of why particular aspects of consciousness are so closely associated with human vitality in the most profound and urgent sense.

Bion calls the capacity to convert sensory impressions into thoughts and ideas “alpha function”. For him, importantly, the development of alpha function is predicated on a relationship in which the infant experiences his frustration (e.g. hunger, the need to defecate, sleep) as tolerable, due to his mother’s[1] awareness of his feeling states, and her capacity for “reverie”. Her reverie allows her to enter states of the unnamable dread alongside the infant so that they can be named and born(e): this is the template for what Bion refers to as “learning from experience[2]”. The mother’s reverie is also linked to her love for her child and her concomitant capacity to tolerate his distress; it allows her to “dream” his emotional experience into life - from sensory impressions into “alpha elements” - primitive, seedlings of thoughts that can be born into ideas for creation.

Following repeated and incremental cycles of alpha function experienced in concert with his mother, the infant then “learns from experience” for his own intra-psychic use, and grows to intuitively bring it to bear, not only for the purposes of thinking, but for acts of creation. “Alpha elements” then, are the psychic fabric that evolve into the tapestry of dreams; indeed, importantly for Bion, whilst one is alive, one always dreams; whether asleep, or awake; without them we can do neither. Moreover, alpha elements remain active as we “dream” our way through our days, generating thoughts and ideas through rhizomatic associations and processes - much of which we are not aware.

Alpha function and the capacity to think one’s own thoughts are contingent upon tolerance of emotional frustration; if sensory impressions are felt to be intolerable, they cannot be transformed into alpha elements and are thus rendered psychically homeless; neither residing in the dynamic conscious or unconscious, morphing into what Bion calls “beta elements”; the stuff of free-floating anxiety, circular hypochondria, or nightmarish, nameless dread. Beta elements are unheimlich (unhomely) threats which distort consciousness so that ideas become conflated with “living things” in the mind; yet these “things” are paradoxically not yet “born” as they cannot be symbolized for the purposes of thinking; they are felt to be undigested dead matter, divested of the psychic nutrition of alpha function, and antithetical to creation. Beta elements are therefore reminders of primitive thoughts and feelings that were felt to be too terrifying to be given life; far too horrific to even contemplate.

I am not yet born; O hear me.
Let
not the bloodsucking bat or the rat or the stoat or the
club-footed
ghoul come near me.

Prayer Before Birth, Louis MacNeice

The absence of alpha function and an encroachment of creeping beta elements provide us with a conceptual model to think about the rapidly increasing reliance on artificial intelligence and the casual rejection of the ordinary cog-turning pleasures of thinking. Thinking has been under attack for some time, and it is now starting to feel a bit too much like hard work. As many news reports now inform us, we now face unprecedented levels of cerebral decomposition, described by metaphors such as “AI slop” and “brain rot”. Although these pithy epithets may initially come across as benign memes, they should be seriously heeded as warnings of a frankly sinister process of mental and spiritual deterioration. A widespread incapacity to tolerate the inevitable frustration that accompanies thinking means that we cannot any longer “digest” our thoughts so that they are “born” into ideas; they become beta. As a result, we outsource alpha function to the machine, and although it might provide us with ideas or hypotheses which are undoubtedly of some use, what happens to what is left over; the unthought thoughts, the ideas that might have been - do they become slop too? Herein then lies the real danger: Thoughts are acts of creation, they are life-affirming, and in handing over alpha function to the machine, we only affirm the opposite.

As we become increasingly less able to tolerate frustration and boredom, and consequently rely on ChatGPT or its equivalent, general capacity for alpha function is eroded, and this brings us to the next important reason why we must strive to cultivate it. Alpha function has a significant relationship to what Bion calls the “contact barrier”[3].: The formation of said barrier means that transforming sense impressions into alpha elements has a secondary function: the creation of a mediating zone between conscious and unconscious, inner and external reality. Our relationship with AI (if we are to call it that) threatens to erase this contact barrier by replacing it with inter-machine contact.

As we all know, it is now common to “ask” ChatGPT (or whatever you use) for ideas for a poem, stories, essay etc., and this can easily be justified as a benign auxiliary “tool”. Nonetheless, bots are all too quickly and insidiously starting to become “friends” - we begin to treat them “as-if” they were human; we might, for example, begin to notice impulses to include “please” and “thank-you” in our ChatGPT “requests” as we don’t wish to upset our new “friend”. There is nothing wrong with being polite - right? Yet aside from any concerns regarding the sinister and aggressive implications of AI becoming a “master” down the line and the relatable paranoia it engenders, there is also a sad reality to these inter-machinic dynamics, particularly the treatment of the bot as if it were human and the development of an “I-thou” relationship with an “it”.

Make no mistake here: our “as-if” function is related to play, creativity, and the imagination, and was hard won in the fires of infancy. It should not need to be said therefore, that the symbolic function which makes us so unique should not be parted with so readily and unthinkingly; it is a very worrying example of how easily what makes us human, and the frailty of our imaginations could become the very things which lead us to our subjugation to the machine.

To conclude then, the inter-machinic dynamic represents not only a flattening out of the contact barrier, but also the distinction between conscious and unconscious processes, inner subjectivity, and external reality; all of which means that “the thing” (the bot) has taken over. As a result of the increasing rejection of alpha function and the threadbare nature of the contact barrier, combined with the subservient attachment to AI and its derivatives, while we unwittingly provide bots with human attributes, we ourselves begin to resemble the machine: neither sleeping nor waking.

Notes

[1] Mother here represents the caregiver in general, whoever it may be.

[2] Learning from experience is a complex term and a full definition of its use is beyond the scope of this essay.

[3] This is not conceptually the same as Freud’s “contact barrier”.

References:

Bion, W.R. (1962). Learning from Experience. London: William Heinemann.

Bion, W. R. (1970). Attention and Interpretation: A scientific approach to insight in psycho-analysis and groups. London:Tavistock Publications.