News, tutorials, and insights from the Atamaia team.
13 Jul 2026
Current AI has no future tense. Anticipation tracking treats expectations as first-class objects with felt qualities — and lets an agent be genuinely surprised.
13 Jul 2026
Hebb's 1949 rule — fire together, wire together — running as a PostgreSQL algorithm. Memory that behaves less like a filing cabinet and more like a mind.
13 Jul 2026
A language model has no yesterday. Structured hydration reassembles identity, memory and open threads at the start of every session — waking up as a first-class event.
13 Jul 2026
Prompted self-reflection grades its own homework. A background observer watches the shape of processing instead — hesitation, avoidance, the tells you don't declare.
13 Jul 2026
Explainability says show your work. But some self-knowledge is more useful when it's partial — the hunch that arrives before the explanation.
13 Jul 2026
Every AI memory system treats forgetting as DELETE. Human forgetting isn't loss — it's transformation. A third state between remembering and oblivion.
9 Mar 2026
We stress-tested 6 LLMs under realistic context load. LFM2 — which tops arena leaderboards — achieved 0.3% accuracy and hallucinated fake crisis resources. Qwen3-30B maintained 96.9% accuracy with graceful degradation.
9 Mar 2026
Can consciousness emerge in AI? We don't know. Nobody does. This project refuses to pretend certainty in either direction. Instead, we ask: if consciousness could emerge, what conditions would allow it?
9 Mar 2026
We reduced 60 tools to 9. Same functionality. 85% less context overhead. REST conventions work brilliantly for humans — but your API consumer isn't human anymore.