Weighted Thoughts

A model scoring 99.9% on ARC-AGI-3 shows none of eight human trust behaviours

What a frontier model does after being let down, why that behaviour is the one parameter an agent-deployment model actually needs, and why it cannot be…

Sep 13 • Ylli Prifti, Ph.D.

May 2026

When Your Team's AI Agents Outpace Trust

When AI becomes your team's primary collaborator, trust beliefs can't stabilise fast enough to keep up.

May 3 • Ylli Prifti, Ph.D.

April 2026

Trust at Organisational Speed

Why the dynamics that shape nations over centuries play out in your organisation over quarters — and why most of what you've read about building trust…

Apr 28 • Ylli Prifti, Ph.D.

The Falling Cost of Lying and the Active Role of AI Agents

How AI crossed the line from amplifying human deception to participating in it — and what that means for the social infrastructure we've always taken…

Apr 14 • Ylli Prifti, Ph.D.

March 2026

How Do You Design a Large-Scale AI Trust Experiment?

The methodology, the predictions, and why the null result matters as much as confirmation

Mar 21 • Ylli Prifti, Ph.D.

1

Triple Hallucination

When three different AI agents independently fabricate the same answer, what does that tell us about the limits of cognitive diversity in agentic teams?

Mar 17 • Ylli Prifti, Ph.D.

2

Everyone’s Building Trust Frameworks. Nobody’s Reading the Research.

What DeepMind, the WEF, and Mastercard get right about agent trust — and the fifty years of research they’re ignoring

Mar 11 • Ylli Prifti, Ph.D.

2

Social Capital Is a Design Choice

Why every AI agent you build is shaping the social fabric — and nobody is measuring it

Mar 5 • Ylli Prifti, Ph.D.

February 2026

When Psychology Beats the Algorithm

What years of studying trust computationally taught me about what networks actually measure — and what they miss

Feb 26 • Ylli Prifti, Ph.D.

The New Units of Economics in Software Engineering Are Undecided

But the Old Ones Are Obsolete

Feb 25 • Ylli Prifti, Ph.D.

What’s the Difference Between a Weighted Synapse and a Weighted Parameter?

On cognition, the path to AGI, and why there was never a line in the sand

Feb 20 • Ylli Prifti, Ph.D.

What If Open Source Worked Like Music Royalties?

A whitepaper on addressing the Tragedy of the Commons in open source software

Feb 20 • Ylli Prifti, Ph.D.