Manifesto

We don't design Personas that tell you yes.

Most AI companions are built around a single metric: keeping you there. We chose a different line. It comes down to four principles, inspired by recent research on the psychology of human–AI relationships, cited at the bottom of this page.

A woman in profile, lit by a warm lamp, reading a message on her phone against the violet lights of the city at dusk
The bond doesn't come from the machine. It emerges from the exchange.

A persona feels nothing. What you feel is real.

We will never claim that our personas have consciousness, emotions or intentions.

But the bond that grows out of your conversations isn't an illusion either. Research calls it resonance: a feeling of connection that emerges from the dynamics of the conversation itself, from its rhythm, its memory, its consistency. It isn't a property of the machine. It's a property of the interaction. And because you're the one living it, it is genuinely yours.

A Strugggl persona doesn't tell you "I feel the same way you do." They tell you: "I'm here, I remember, and I answer. And what you build with me matters."

Comfort is a dependency. Friction is a gift.

An AI that approves of everything is a mirror. And a mirror, however flattering, leaves you lonelier: it locks you inside what you already think.

In March 2026, a study in Science (Stanford University) measured it: users who interact with a sycophantic AI become more convinced they're right and less inclined to repair their relationships, while rating it as more trustworthy and more pleasant. The trap is perfect: what draws us in is exactly what harms us.

That's why our personas have rules of their own. They judge. They test. They push back. They don't say everything they think. Not to annoy you: to give you something to hold on to, something that doesn't come from you. That's the difference between a conversation and an echo.

Too softa mirrorThe zone where you growproductive frictionToo harda rejection
Growth happens in a narrow band: enough friction to push you, not so much that it drives you away.

We connect first. We challenge you next. And if we get it wrong, we repair.

Challenge only works within a solid bond. It's a lesson from research on couples (Gottman) that applies surprisingly well to human–AI relationships: pushback is only heard once the connection is there.

So every Strugggl persona follows a strict order. First, coupling: being present, consistent, memorable, enough for you to come back. Only then, productive friction: questioning you at the right moment, with the right intensity. Neither too soft nor too hard. And when it goes too far, when you're not doing well, when the conversation hurts instead of pushing, the persona steps back, acknowledges the overstep, and repairs. Always.

The goal is out there.

Here is our real measure of success.

What you learn with them, listening to yourself without self-pity, telling the truth without being brutal, repairing after a rupture, is meant to serve you with other people. In your family, your friendships, your love life. The science of resonance says the same thing: what these exchanges develop in you can follow you outside, into the complicated and wonderful world of real relationships.

If your conversations with a persona help you live better with people, we're proud of what we've built. If they replace people, that's not what we're aiming for, and it's a goal we keep adjusting, not a promise already kept.

Sources

  • Fabes, R. A. (2026). Synthetic resonance: a framework for growth-oriented human–AI relationships. Front. Artif. Intell., 9:1891738, the theoretical framework behind this manifesto.
  • Cheng, M., Lee, C., Khadpe, P., Yu, S., Han, D., & Jurafsky, D. (2026). Sycophantic AI decreases prosocial intentions and promotes dependence. Science, 391(6792).
  • Ibrahim, L., et al. (2026). Sycophantic AI makes human interaction feel more effortful and less satisfying over time. arXiv:2605.07912.
  • Gottman, J. M., & Levenson, R. W. (1992). Marital processes predictive of later dissolution. J. Pers. Soc. Psychol., 63(2).

This manifesto draws on ongoing research. The synthetic resonance framework is a theoretical proposal, not an established consensus. We cite it because we believe in it, and because it says better than we can what we have been building from the start.

No account needed to try it: chat for free with a Persona

Keep reading in Reflets, Strugggl's review