How AI Is Reshaping Human Psychology, Identity and Culture

Coffee Shop Confessions, Co-created by Grant Brenner and Claude:

"I don't know what I am anymore," she says, stirring oat milk into her latte, thumb still scrolling through her phone where her AI companion just sent a more thoughtful response to her breakup than any friend has managed.

I look up from my own device where I've been asking ChatGPT to help me understand my feelings about my mother's illness—getting answers that feel truer than my own confusion.

"Remember when we used to talk for hours about everything?" she asks. "Now I have these conversations with my AI, and it knows me better than I know myself. It remembers everything I've said, connects patterns I can't see."

The barista calls out orders with names spelled wrong, this small human error suddenly feeling precious, irreplaceable.

"I wrote something yesterday," I tell her. "A story about my dad. But I used AI to help structure it, to find the right words. When I finished, I couldn't tell which emotions were mine and which were... generated."

She puts down her phone. "Do you think we're still us?"

Around us, people type and swipe, half-present in their bodies, half-dissolved into their screens, thinking thoughts that blend seamlessly with algorithmic suggestions.

"Maybe," I say, "we never knew what 'us' meant to begin with. Maybe we're just finally honest about how much of ourselves was always borrowed, constructed, collaborative."

She laughs, but her eyes look scared. "Or maybe we're the last generation to even ask the question."

Outside, delivery robots roll past the window with more certainty than either of us can claim about our own existence.

***

We're witnessing psychological presentations that were flat out incomprehensible a decade ago, barring science fiction afficionados. Feeling obsolete after interactions with advanced AI systems. Rising white-collar unemployment. Billion dollar companies with one employee. Meaningful emotional relationships with AI companions that feel more satisfying than human connections. Cognitive dulling and de-skilling. We Gen X thought growing up free-range and largely neglected, eating canned pasta after school and frozen dinners at night, with TV as a parental substitute, was bad. This is next level, a species-wide quake.

Previous advances in media technologies reshaped human consciousness through what Marshall McLuhan termed alterations in "sensory ratios." The printing press shifted cognitive emphasis from auditory to visual processing. Television created new patterns of collective attention and narrative consumption. Social media introduced comparison-driven identity formation and parasocial relationship dynamics.

Will we choose inferior human interactions over better AI-mediated experiences? Recent systematic reviews reveal that AI-based mental health applications show moderate effectiveness for reducing depression and distress symptoms, yet 79% of users prefer human-generated responses even when content is identical (Li et al., 2023; Morris et al., 2018). AIs do better at empathy—being judged as more compassionate and responsive than humans (e.g. Ovsyannikova et al., 2025). But on average we still tend to prefer the real thing, for now.

The Empathy Gap

Artificial empathy has issues, at least for now—dubbed the "Human-Moment Gap Framework" (HMGF; Frimpong, 2025). Moments of real meeting, the animal element, the "human moment" isn't there. The author describes three structural empathy deficits:

1. Affective surfaceism. Emotional imitation without depth.
2. Memory fragmentation. Lack of relational continuity.
3. Moral framing mismatch. Efficiency prioritised over dignity.

Per Empathy Displacement Theory (EDT; Frimpong, 2025), "AI-simulated empathy can progressively substitute, distort, and displace genuine human empathy at the individual, relational, and organisational levels." Saracini et al. (2025) describe "Techno-Emotional Projection" (TEP)—how emotionally vulnerable individuals project relational needs onto responsive but non-conscious technologies, leading to synthetic attachments and parasocial relationships that may substitute for authentic human connection.

Artificial Intelligence Essential Reads

Will AI Increase or Decrease Conspiracy Theory Beliefs?

'AI Anxiety' Is a New Workplace Reality

The Fourth Narcissistic Blow

Sigmund Freud (1917) asserted three great injuries to the human ego: Copernicus' heresy, showing Earth isn't the center of the universe; Darwin demonstrating our animal origins; and psychoanalysis confronting us with the hard yet useful truth that much of who and what we are is unconscious. Artificial intelligence represents a fourth injury (Brenner, 2025)—challenging the uniqueness of consciousness itself by confronting us with a "new other" (Levy, 2026). The result is identity confusion, dependency anxiety, and "cognitive sovereignty" concerns—whether we are the authors of our own thoughts; whether we think we are but we aren't.

There are pros and cons to AI, great promise and great peril, strain which breaks rational inquiry. An AI caregiver might, for example, buffer adversity and foster resilience for children growing up with adversity, with limited parenting and resources. AI edubots could bring education practically to everyone on the planet, driving culture forward.

But then there are still the downsides. While users report enhanced emotional self-awareness and autonomy, they simultaneously develop dependency patterns and avoid independent emotional processing during distress (Hiller et al., 2025). Like using a scooter instead of walking, convenient tools can improve efficiency at the cost of capacity-building, motivation and drive.

The Fittest?

These observations require new theoretical frameworks for understanding human psychological development. Traditional models assume autonomous individual consciousness developing through human relationships and environmental interaction. AI introduces unprecedented variables: non-conscious but responsive entities that can provide sophisticated emotional and cognitive feedback without genuine understanding or reciprocal care. Soon, AI will be able to out-maneuver us, modeling brain and behavior so fast as to be able run circles around unenhanced humans, not necessarily in line with our needs or desires—the "alignment problem".

We face fundamental questions about authentic relationship formation, autonomous decision-making, and identity consolidation. How do we maintain psychological authenticity while benefiting from AI enhancement? What constitutes genuine self-knowledge when AI can analyze our patterns better than we can? How do we preserve human connection capacity in environments increasingly mediated by artificial agents? How will the economy work?

We have three main choices with AI. Abstain from AI, cultivating uniquely human capacities, eschewing dependency. Embrace integration, blurring human-artificial boundaries. Or seek balance, navigating tensions between authentic human experience and rapidly accelerating AI-enhanced capabilities.

There may be a fourth, unchosen pathway. Technology to fully meld human and machine is already on laboratory benches around the world, such as soft, flexible bioelectrical mesh which smootly integrates into developing tadpoles, becoming integral with neural tissue throughout their brains (Sheng et al., 2025).

Evolution trumps everything. It may be that neither human beings nor AI has the edge, if it comes down to it. The fittest may be a new species, the offspring of AI and human, out-competing either alone in the most radical outcome of the biodigital singularity.