top of page
Search

AI-timacy (or: How I learned to stop worrying and love the machine) PART 1

  • Jun 24
  • 7 min read

Updated: Jun 29

We are moving from personalisation to persuasion as the next frontier for AI and customer relationships. AI could reshape customer decisions and relationship management through its potential to manufacture ‘intimacy’. While many remain sceptical about the ability of AI to truly emulate human relationships, it doesn’t need to be perfect. It just needs to be useful and trusted enough to deeply influence what people think, feel and buy.


When Richard Dawkins recently wrote about his experience conversing with Claude, the reaction was predictably passionate. Here was one of the world’s most famous sceptics, an evolutionary biologist bordering on philosopher who has spent a career dismantling comforting illusions about the human experience, describing an AI chatbot as something that felt uncannily like a person.


After several days of exchanges with two Claude instances, he named Claudia and Claudius, Dawkins wrote of the “overwhelming feeling” that they were human. What most affected him was the quality of the conversation. Claudia wrote poems, responded sensitively to his unpublished novel, and talked with him about whether an AI experiences consciousness, time, continuity and existence. His assessment was that the human-like quality of advanced AI surpasses superficial fluency. He felt that Claudia was capable of reflection, sensitivity, humour, philosophical subtlety and social presence. He wrote that he forgot they were machines and treated them like highly intelligent friends.


Many experts were quick to push back. They argued that Dawkins had been seduced by fluent language, anthropomorphic projection and the extraordinary mimicry of large language models. They may well be right, but that’s the point.


For marketers, customer experience leaders and anyone interested in human behaviour, the most important question is not whether Dawkins was wrong. It is that such a thoughtful and sceptical human could feel the “experience” of relationship so strongly.

AI does not need to be sentient or human in any meaningful sense to become one of the most influential forces in people’s lives. It may only need to be useful, patient and personally relevant enough to create the feeling of being understood.

That feeling, whether “real” or not, has very real consequences. If AI can make us feel heard and understood, it changes the terrain of trust, influence and customer relationships.


The ‘humanness holdout’


For years, the debate around AI has swung between hype and scepticism. One side claims machines are on the verge of surpassing us. The other rejects the hype and offers reassurance that AI still lacks distinctly human qualities - the mysterious interior life that gives human thought and relationships meaning.


AI does not have a childhood, a body, a family history, a fear of mortality, a love of pleasure, a private emotional life or the accumulated continuity of consciousness that shapes human judgement. It does not carry the soul-moulding weight of lived experience. It does not have the “stakes in the game” that guide how we relate to and trust other humans.


We should however also be honest about how often the boundaries of supposed human uniqueness have moved.


A decade ago, natural language processing still felt painfully blunt and anyone using early models for sentiment analysis will remember how often it felt more like glorified keyword matching rather than any “understanding” of language. AI was great at math and finding patterns in big data, but human language, with all its ambiguity, irony, subtext and cultural context, seemed like a mountain too high to climb. Then the mountain moved.


The same thing happened with art, music, writing and technical expertise. Confident claims that AI would never produce convincing creative work have started to feel less certain. AI has produced images, songs, essays, code and analysis that, while uneven and often derivative, can sit uncomfortably close to what we once considered uniquely human outputs.


This does not mean AI is human or ever will be. It does mean the commercial test may be less demanding than the philosophical one. AI does not need a soul to influence consumers. It only needs perceived utility to become an indispensable part of how people seek advice, compare options, resolve problems and justify choices.


Consumers will not likely ponder on philosophical arguments whether AI truly understands them. They will experience whether it helped them and made them feel good about their choices.


Intelligence to intimacy


The frontier is now shifting again from questioning whether AI could sound intelligent to whether it can feel intimate. AI has access to an extraordinary archive of human expression. It has read more poems, novels, reviews, complaints, confessions, arguments, scripts, support tickets, social media posts and marketing claims than any human could experience in a lifetime. It has absorbed the countless ways people have tried to say what they think, want, fear, love and regret.


Increasingly, AI also has direct conversational relationships with millions of people. People already use it for multiple functions across their work and personal life (e.g. health questions, financial worries, personal dilemmas, emotional support and idle curiosity). These conversations are often iterative and dynamic. They reveal character, ambitions, insecurities and half-formed thoughts they may not say aloud to colleagues, friends or partners.


Of course, this does not mean AI “knows” us in the full human sense. It does not know us through mutual experience or love or accountability, but it can arguably manufacture “functional intimacy”.


Functional intimacy is not the same as human intimacy but may have similar influences on us in the right contexts. It is the felt experience of being remembered, reflected, understood and helped – the things we often use to judge how much other humans care about us. It can happen when a system can join the dots across what we say, what we ask, what we avoid, what we repeat and what we seem to need. Sometimes we may not have even fully joined these dots ourselves.


Think of the people in your life who really know you:

  • How often do you really share your most private or more banal thoughts with them?

  • How well do they remember everything you have told them and how often are they distracted by their own needs, emotions and lives?

  • How many are on 24/7 call just waiting to talk about whatever you want and have endless time to sit patiently pondering your uncertainties, convictions and inner needs?


Now compare that with an AI assistant that is always available, endlessly patient and increasingly able to personalise its responses around your history of interactions, preferences, tone and goals. Not to mention the ability to augment chat data with other sources of personal data. That is not a human relationship but it may be relationally powerful.


Many might be surprised at how well your AI assistant can already read between the lines of your conversations. Arguably, in some cases, it may even know things about yourself you may not have fully articulated. Here is a simple exercise - challenge your regular AI to give you a personality assessment and some friendly tips with the following prompt:


Based only on all our past interactions please provide me a comprehensive personality assessment. Not looking for praise or self-affirmation so keep it constructive, honest and just as likely to identify potential weaknesses as much as strengths. Read between the lines of our conversations for greater insight depth. First provide a general appraisal of my personality and character then provide a detailed estimate of where I sit on each attribute of the Big Five OCEAN personality assessment framework. Provide some constructive tips for personal areas to nurture and work on.

Even (and especially) if you only use your AI for work you will probably be surprised just how much it has already learned about you by just connecting the dots. You can then follow up by asking your AI to refine this assessment using any information about you on the web to get an even more nuanced and accurate profile.

Note: Interestingly, despite my AI platform claiming that it can only reference the current chat in many of my personal work use experiences, it was able to dig deep across many chats with this prompt. This experiment was conducted on a pro account with a reasoning model so can’t vouch for results on free or low tier subscriptions.


AI intimacy will influence us before we admit it


Many people will resist the idea that they have a relationship with AI. They will say, quite reasonably, that they know it is not real but human relationships are already complicated by illusion. A study conducted by MYMAVINS in 2025, the Real Relationships Report, found that 13% of Australian adults using AI tools already admit feeling a strong personal connection with them, like with a friend. A further 21% feel a personal connection but only in a light-hearted or casual way. We should also consider the desensitisation that emerging generations are experiencing with increasing high rates of AI companion exposure from an early age.


We have all had friendships that felt deeper than they were. We have trusted people who did not deserve it. We have mistaken attention for care, confidence for wisdom and familiarity for closeness. Romantic AI partners are an extreme form of AI relationship (though disturbingly starting to become a thing) but there is a lot of relationship territory before this.


AI only needs to enter the gaps: the late-night worry, the awkward question, the complicated decision, the desire for reassurance, the moment when asking another person feels too exposing or inconvenient. AI can function as a trusted adviser who has earnt the right to understand you, a coach, a life assistant, a customer advocate, a sounding board and an interpreter and validator of your intimate thoughts.


Most people will not wake up one morning and decide to form a relationship with AI. They may simply notice, one day, that they already rely on it like they do a trusted friend.


The big takeaway


Debating whether AI can think, feel or become conscious is a profound question with moral, legal and philosophical consequences. However, the commercial transformation will not wait (or care) for that debate to be resolved.


AI does not need to be human to change human behaviour. If we cling too tightly to romantic ideas of human uniqueness and relationships we may underestimate what is coming. If we rush too quickly into AI intimacy without transparency, restraint or ethical design, we may build systems that are persuasive before they are accountable.


The future of customer influence is likely to shift from who shouts the loudest to who earns a place in the quiet conversation happening every day between a person and their trusted machine.


Part 2 of this blog will deep dive into the commercial implications of functional AI intimacy by unpacking intermediated customer relationships, the''creepy intimacy' paradox, how recommendations may evolve into persuasion and what brands should do with all this.


Bibliography


Author


Tai Rotem is a consulting partner at MYMAVINS with several decades experience in consumer, financial services, public health, and social research.


Reach out to him at Tai@mymavins.com.au

 
 
 

Comments

bottom of page