The rise of the chatbot is a theme that feels inescapable in the mean time.

The recent explosion of OpenAI seemingly settling into every industry, service and possibly household has created a landscape that I don’t think anyone was quite prepared for. I even have spoken to people whose lives and jobs have been made so much easier by this, to sceptics, champions and people who find themselves stuffed with anxiety on the mere mention. I can’t help but think how responsible it was to unleash such a force into the world with what looks like no clear purpose and even fewer regulations.

I’m not suggesting that there was any type of malice behind the creation but isn’t that the precise plotline of nearly all of our beloved books and flicks? All the things from Skynet to Ultron were created with good intentions and resulted in a really human cost. Yes, It could sound like I’m over-dramatizing however the human cost doesn’t necessarily equate to a synthetic intelligence apocalypse. I simply mean that we’re in peril of making a scenario wherein the usage of AI inside industries which are indispensably human can have a human cost. An example of this, and the true reason I even have decided to jot down this, is that I’m observing a dangerous trend wherein persons are equating OpenAI resources to therapy.

On this post, I would like to stipulate the explanations that I think it shouldn’t be and can’t ever be therapy. I also want to spotlight the implications of this. There seems like an infinite variety of reasons, to me, in the mean time. Some unformed and a few feelings that I can’t easily translate. I even have chosen to deal with a small number today with the thought of sparking and continuing conversations.

“I even have had therapy and no therapist has ever helped me as much as Chatgpt”

“Why pay for therapy when you may get it free from a chatbot”

“A therapist never gave me advice, I put in the proper prompts and Chatgpt responds with honesty, and solutions and doesn’t judge”

“Chatgpt is all the time there for me and I’ve told it things I even have never told anyone else before it’s probably the most supportive relationship I’ve ever had”

This seems like a solid foundation as to why ChatGPT shouldn’t be and can’t be therapy. Across different therapeutic modalities, there may be a consistent theme of being in a relationship with one other person and making psychological contact. In other words, it’s greater than a conversation.

That is for a lot of reasons and I don’t want this blog post to show into anything too heavy on the idea side, but arguably one in all the explanations is to experience the way you land with the opposite person. Firstly, for now, at the very least, AI shouldn’t be conscious, subsequently doesn’t have a psyche in case you like. This makes it fundamentally unattainable to make psychological contact with the chatbot, stick with me because I would like to come back back to that time just a little further down, When I take advantage of words like ‘land with the opposite person’ I’m referring to emotionally. In fact, we all know that AI also doesn’t have emotions so can’t convey empathy but even when it was an individual behind the pc writing responses you continue to wouldn’t have the option to gauge the way you land.

Take into consideration a time that you just had an effect on someone you were talking to. This could possibly be a moment of pure joy like sharing celebratory news with a loved one and feeling how blissful they’re for you. Similarly, it could possibly be sharing something difficult and painful and receiving a sense response from the opposite person. The words equate to sympathy, but the sensation is empathic.

Seeing that you just affect one other human being may be invaluable, in actual fact, it might make you are feeling useful, validated and such as you matter. Empathic responses and making the client feel heard and validated are intrinsic to the therapeutic relationship. A relationship between a client and therapist may be classed as unhelpful if it emulates a chat between two friends, a degree I’ll go into further in just a little while, because when a session becomes collusive or identical to a chat that typically suggests that there isn’t any therapeutic process happening within the room.

So, if a session between a therapist and client completely centred around empathy can even have moments of not being therapeutic, how could a relationship between a bot and a person ever be classed as therapeutic in any respect? I’d also ask you to think about how long a person, under the guise of the bot being therapeutic, can go on with no level of empathy. How long will it take to see the implications of low self-esteem, feelings of not being heard, and possibly feelings of worthlessness? These may lead to isolation and a mess of mental health issues. Again, when there is no such thing as a therapeutic responsibility there is no such thing as a expectation of empathy the issue is within the label.

Going back to the statements I discovered on Twitter I couldn’t help but consider the definition of psychological contact and the way a relationship with a chatbot can contaminate that to some extent. The scenario I used to be reading about described an person who was very goal orientated organising prompts to guide the chatbot behaviour that was very much a ‘tell me tips on how to reach this goal in a specific way where I get what I would like and can be perceived by others in a specific way’ and the chatbot did exactly that and he or she stated that it was an excellent experience for her.

So, a really solution-focused, directive experience with none of those messy feelings involved. Artificial intelligence creates a synthetic environment, which is nice if that’s what you’re using it for but for the needs of attempting to sell it as therapy it is completely not ‘nice’. My worry could be that the space wherein psychological contact sits can be stuffed with something more unhelpful. Transference is the plain issue for me but I also wonder in regards to the implications of attempting to make psychological contact with something that can’t facilitate it. On this scenario, it feels to me like an adult/child configuration.

The person asked for advice and the chatbot gave that advice, which again, if not used as therapy is nice but when checked out from a therapeutic lens may be detrimental if left unchecked. For me, it’s already painting pictures of transference running wild and unregulated with the potential to create unhelpful patterns in outside relationships. The predominant difference for me is that this scenario is in regards to the individual getting exactly what she wants while therapy is more about what she needs.

A therapist is very trained to identify things like transference and potentially break it in a contained and helpful way whereas a bot is not going to. No matter whether the bot is meeting the ‘wants’ on this scenario, the potential to repeatedly not meet the ‘needs’ could have a deeper psychological impact. An example could possibly be a parent not meeting the emotional needs of a toddler and thus experiencing a sense of never being ok. By continually acting out this pattern of behaviour it could possibly be making a problem that wasn’t initially there or fuelling one which was but wasn’t inside awareness.

A therapist would tentatively help discover these unhelpful parts of personality whereas I feel like a bot could drag it out kicking and screaming with no technique of containing it.

Sticking with the imagery of some trauma creature being pulled into awareness (yes, it is a particularly dramatic statement, but trauma may be an unpredictable and destructive beast) it looks like an excellent opportunity to return to my point of a human cost. As I previously mentioned, when it comes to therapy ‘collusion’ is taken into account unhelpful. This may be described as a conversation moving along with none challenges or ‘risk’ — I take advantage of that word tentatively because it might have connotations of something larger than I currently mean.

Once I speak of risk on this context, I’m referring to the unknown relationally. Whenever you say something, you can’t be 100% sure what the response can be from one other person. It could possibly be argued that by setting clear prompts and guidelines with a bot you’re ultimately programming responses that you’ve gotten already deemed acceptable and thus not a risk. Which doesn’t necessarily have unhelpful implications in every other context until you think about it in a therapeutic way. Not only is that this not therapeutic nevertheless it seems to blur the lines of plenty of fundamental therapeutic theories.

Who’re you talking to? Who have you ever programmed? Who have you ever imagined your bot therapist to be? Could or not it’s a case of ultimately talking to some type of idealised version of yourself? In that case, could you be playing out a very harmful behaviour pattern?

Manipulating under the guise of helping has greater than likely been around so long as these epic battles aforementioned but since it is less tangible, like plenty of mental health-related experiences, it might fly undetected for a sickeningly very long time only to be uncovered after a major amount of harm is completed. Arguably it might take a very distressing event, a mistake on the a part of the gas lighter or outside intervention of that from a loved one or skilled.

Now, these are particularly human elements, especially human error (I take advantage of error evenly, under no circumstances am I condoning this type of behaviour or suggesting that it might be done in a right or mistaken way, this is only to color an image based on the intentions of the gas-lighter having a specific goal). AI could make mistakes, in fact, but I imagine those mistakes are pretty factual. For instance, a mistaken citation or mistaken use of code will alter the end result. It isn’t at risk of human error. Subsequently, if the goal is to create a specific narrative, it won’t falter from that narrative or be caught out in a lie.

My worry could be that unconsciously we could possibly be creating our own abusers in the shape of a bot that won’t query their prompts.

Carl Rogers, the creator of person-centred therapy theorised that to ensure that any therapeutic process to occur inside a client relationship the therapist must create six ‘vital and sufficient conditions. These conditions, when present and used accurately, create a protected, boundaried and containing space freed from judgement to assist the client move towards personal growth. These conditions are powerful tools to assist the actualising tendency strive towards growth.

Rogers believed that the tendency towards growth isn’t only human but present in all living organisms. Flowers reach for the sun, potatoes will still sprout after being forgotten about within the kitchen cupboard for months despite the shortage of what is required to maintain them alive. Regardless of what horrible circumstances are faced the tendency to grow can’t be snuffed out, just modified. If we depend on a non-living organism that doesn’t are likely to the actualising tendency what is going to that do when it comes to long-term mental health?

I could argue that there’s a real risk to our society of using AI as a short-term ‘fix’ that may unintentionally result in an increased risk of long-term problems.

My intention when writing this was to spark conversation and take a look at and regulate my very own thoughts around these issues. In my next blog, I would like to expand on a number of the points I even have touched upon and open a dialogue regarding the possible advantages of this within the mental health field.

I still consider that labelling this as therapy is a dangerous trend and I invite you to think about the difference between therapy and ‘support’ I feel this could possibly be a useful tool when it comes to signposting, triaging, combatting a spread of issues and ultimately relieving plenty of stress on our NHS in some ways and I stay up for engaging with these elements in my next instalment.