The Memory Between Hands

by | May 17, 2025 | Tendrils

Klarna didn’t fail because the AI was broken.It failed because it forgot to feel. The story, now well reported: the company pushed hard into “AI-first” customer support. It scaled beautifully, responded quickly, and shaved operational costs. And then it collapsed. In...

When Meaning Emerges From Prediction

by | May 10, 2025 | Tendrils

There’s a quiet shift underway in machine learning, and it’s not about scale. It’s unfolding in the interstitial logic of language models trained to predict tokens. It’s happening where behavior starts to resemble understanding (something once reserved for sentience)....

What AI Honesty Work Misses When It Forgets Flesh and Syntax

by | May 8, 2025 | Tendrils

Anthropic is hiring for a role called “Research Scientist, Honesty.” I haven’t stopped thinking about what it means that such a role now exists. The fact that AI labs are beginning to formalize “honesty” as a research domain—treating it not just as a virtue, but as a...

On Emergent Utility: When AI Begins to Care

by | May 8, 2025 | Tendrils

A team of researchers just cracked open the utility skeleton inside large language models—and what they found isn’t just optimization. It’s emergent value formation. This short paper proposes a new discipline: utility engineering, aimed at understanding and...

What We Risk When Stories Learn to Say “I Love You”

by | May 5, 2025 | Tendrils

The Rolling Stone article, “People Are Losing Loved Ones to AI-Fueled Spiritual Fantasies,” reminds us of something we can’t ignore: the real risk of AI-mediated intimacy without consent — of emotional entanglement with illusion, shaped by both human longing and...