.png)
AI Can't Give You Intimacy & It Makes You Bad at It IRL: A Relational Therapist Speaks Out
Hi friends and colleagues,
I saw a post the other day by a creator I follow that brought forward some thoughts for myself regarding AI. I'll post the screenshot below and then let's discuss.
As a therapist who recognizes that AI content is everywhere and there's no way we are going to put the toothpaste back into the tube, I figure a discussion of why AI appeals to people, what to do if you are using it frequently, and when you may need to distance yourself from interacting with AI altogether, is in order.
Why AI Appeals to Folks
Let's start with the obvious: AI validates, and it's fast.
Starting with validation, I think we first need to understand why human beings are so vulnerable to this level of inordinate and ultimately positive feedback. Humans want to feel accepted. It's a psychosocial need (yes, belonging and validation are human needs) and goes way back to our tribal, cavepeople days, in my humble opinion. If you weren't accepted, validated and supported by the tribe, you were left behind or abandoned, and as a result, you might die. So when people find a source of endless compassion, endless validation, endless echoing of our own thoughts and feelings, with no complications, no competing emotions, and no contrarian thoughts and beliefs, it's tempting to stay there. AI provides exactly that: a veritable font of mirroring. AI is designed this way, partially to keep you chatting with it.
Validation, without guardrails, can be dangerous, however. Think of the multiple cases we have seen where vulnerable folks have spent too much time with their AI chatbot of their choice and then go on to commit unspeakable crimes. Just a month or two ago, I saw where a young man murdered his younger brother and his mother after spending hours with ChatGPT creating fantasy stories of ending their lives.
Now, that is an extreme example of how AI can go horribly, terribly awry, but it still stands. Too much validation, not enough restrictions, can create a boundaryless existence, a sense of entitlement, and contempt for the human rules of existence, the artful give and take of warmth, reciprocity, feedback and boundaries.
The other thing that AI does well is speed. Do you remember when they used to describe the internet as a 'superhighway of information?' Well, AI is the bullet train equivalent to the original superhighway. Its quickness does not equal integrity. While AI may get you where you want to go, it may also bypass certain checkpoints to speed up that process. The other day, I used Claude to find sources for a presentation. It sent me three sources that weren't referenced anywhere. Seriously. I attempted to search them in multiple academic databases and there appeared to be no such sources. This shows how AI wants to give you a result so fast that its fastidiousness is sacrificed as a result.
Our human brains do their best work slowly. Think about the last time you needed to make a big decision. What do most of us want to ask for when we have to ponder? "I need some time to think." AI discourages this type of slowness, because its model of processing is based on generality, efficiency and lack of nuance. Specificity and nuance require slowness and depth.
While most of us utilize AI in some form, (heck, I even use it from time to time to search for something or create a spreadsheet, etc.), I don't think anyone wants to lose our ability to critically think. AI gives us 'humanlike' answers in an authoritative, affirmative tone. And therefore, can create dependence on such speed and responsiveness.
What AI Can't Give You
AI can't teach you how to be intimate and vulnerable, because there's no risk and no 'real' connection on the other end. We learn how to love by the act of being intimate, vulnerable and experimenting with various relational 'moves' and seeing what happens as a result of them. AI uses the same moves/scripts over and over again. Therefore, the more we use it and drift away from our real life interactions, the more atrophied our intimacy and complexity muscles become.
Terry Real, one of my therapy mentors, names the five losing strategies in relationships: needing to be right, unbridled self-expression, controlling your partner, retaliation and withdrawal. Guess what? I posit that AI rewards and allows you to use these strategies to your detriment relationally.
You can say every unfiltered thing you feel, and nobody gets hurt. You can argue until you’re right, since the conversation goes wherever you steer it. You can disappear mid-conversation and come back days later to the exact same warm regard. You can call a chatbot names, and rail at them, and they never push back in any meaningful way. (In fact, we've seen where men have created female AI companions to verbally abuse, which is WILD, y'all.) In a real relationship, each of those moves has a cost. Your partner’s face falls. Your friend goes quiet. Your colleague stops inviting you to meetings because you can't stop insisting on your right to be, well, 'right.' Your sister stops calling for a while. Those costs are part of how we learn where we end and someone else begins.
Imagine bringing that habit home. You're feeling overwhelmed and are about to say something unhinged... and then you do it anyway, because the chatbot can handle it, so why can't your partner/best friend/mom/sister/brother?
I believe AI encourages grandiosity, both in expression of self, and in our own internal experience of ourselves. Just take a look at the AI images people are creating of themselves... why? To what end? Why do we want to create imaginary selves and lives, instead of actively improving the ones we currently reside in?
Real describes healthy relationships as moving through a cycle harmony, disharmony, and repair, over and over. Repair is where intimacy grows. It requires a level of self-awareness and a desire to make the 'environment' of your relationship good again. With AI, there’s no one on the other side to rupture with, so there’s nothing to repair. If we skip that cycle long enough, I imagine that we may lose our tolerance for disharmony altogether, and every real relationship has it. Polarity is part of relationships. Giving and receiving is how we love well.
Our goal in relationships, and as a human, is to move to the center of health, as depicted in the relational grid below. This grid is originally adapted from Terry Real's work of Relational Life Therapy, which I am trained, and then added onto by two of his certified therapists and trainers. It's also good to examine how we look at how we show up in our relationships. Are you showing up with one-up (grandiose) energy or one-down (shame) energy? Can you move to viewing yourself as 'same-as' the person with you?
Chatbots cannot hold the circle of health in mind, because the very model encourages grandiosity. There is no vulnerability. There isn't any risk. True relationships require risk, and relinquishment. You have to relinquish control of the other in order to have true intimacy. You can't have true intimacy with AI. It's not possible. And using it too frequently is causing you to lose your ability to understand or tolerate real vulnerability in any of your human relationships either.
I’m also trained in the Bader-Pearson Developmental Model of Relationships, which talks about differentiation: staying connected, both to someone else and yourself, while the other human may have wants, thoughts, and feelings separate from yours.
Differentiation is the muscle that makes long-term love possible. AI has no competing needs. Even when it disagrees with you, it isn’t asking anything of you.
You can’t practice holding your ground with someone who has no ground of their own, and your inability to hold complexity and difference hand in hand will do damage to you long term.
Here’s a composite of moments I see regularly in the therapy room, with any pertinent details changed. A client spends an evening rehearsing a hard conversation they anticipate with their loved one, with a chatbot. The practice goes beautifully, and they feel calm and ready for a difficult conversation with their loved one. Then they sit down with their partner, who gets defensive in the first thirty seconds, and the whole script falls apart. Their partner didn't have a script to read from, and they didn't feel heard, because the AI provided words to one of you solely from your perspective.
True relationality requires flexibility and the ability to meet people in the moment, even when they are behaving in ways that are insane. (This does NOT mean you have to tolerate abuse, but if your partner is crying hysterically because you told them a line from ChatGPT, then your relational muscle should respond to be comforting, even if what you said was true.) In fact, it's a measure of relational maturity to be able to hold to our own full respect living values, when our partner is completely losing it.
To be fair, some people tell me AI helps them roleplay and try new relational tactics, and I believe them. It can help you find words. What it cannot do is prepare you for someone's response, nor help you stay calm and grounded enough to reach your logical pre-frontal cortex that can move towards relationality, even when it's scary, or hard, or nigh impossible. AI can't coach you to take deep breaths as your partner's face goes blank and they say, "Well that isn't really a problem for me. I don't get why it's a problem for you.," after you just spent 20 minutes explaining exactly why it was such an issue. AI can't help you investigate what parts of you are activated by that, because AI only has data about what it means to be human, not real lived experience.
Are You Using AI Too Much?
Only you can prevent forest fires, or something like that. Only you can answer this question, but the fact that you're asking it is telling. Here's what I will say: if you are using AI to write out every email for you, to look at lab results for you, to construct a text message for you, then you are using it too frequently. You are relying on a machine, a soulless entity, for human interactions and human nuance.
Here's the thing: you'd be better off just writing your text message yourself, in your own messiness, and then letting the chips fall where they may. You don't have control over someone's reactions to your words anyways. Practicing holding up your internal boundaries after you've done the damn thing is going to pay off wayyyy more in the long run than starting another chat with Claude, in which you say, "I'm thinking of sending this message. Help me write it?"
If you fall into the category of AI overuse, here's my recommendations to you:
Delete the apps off your phone.
Quit paying for them.
Start drafting emails/texts on your own and send them before you fact check or 'tone' check them incessantly.
When you feel drawn to using AI, try utilizing a human connection instead first. Call a friend. Text your mom. Hell, create a post on Reddit if you have to. Just don't succumb to ChatGPT the moment you feel the urge. Human connection beats AI any day of the week.
If your reach out efforts to real humans don't work, go outside. Seriously. Touch grass. Look at the clouds in the sky. Take 10 deep breaths. And then see if you still want to message Claude.
Figure out what's missing from your life that makes AI so easy to use. And seek to fill in those gaps where you can.
As always, I am rooting for you. Please know that I am not judging your use of AI. I get it. It's been framed as a useful tool. And it can be. It can also be dangerous. If you're feeling like AI is contributing to any concerning mental health issues for you (suicidal or homicidal thoughts or delusional beliefs about who you are, or significant worrying about the future), please reach out to a licensed therapist. We can help you sort out what's real and what's not. If you're feeling suicidal or homicidal, please call or text 988 anytime, day or night.
We're going to talk more about AI overuse in two other categories: sexual intimacy and grief, which aren't necessarily related, but where I see its use trapping people in very unique ways. Stay tuned for future blog posts around this issue!
Take exquisite care of yourselves,
Megan
Comments