Guardat en:
| Autor principal: | |
|---|---|
| Format: | Recurso digital |
| Idioma: | |
| Publicat: |
Zenodo
2026
|
| Matèries: | |
| Accés en línia: | https://doi.org/10.5281/zenodo.19788392 |
| Etiquetes: |
Afegir etiqueta
Sense etiquetes, Sigues el primer a etiquetar aquest registre!
|
Taula de continguts:
- This paper applies Hypothesis-Driven Development (HDD) -- a lean approach where every architectural idea is expressed as falsifiable hypotheses with measurable targets before implementation -- to evaluate predictive processing as a cognitive enhancement for neurosymbolic AI agents. Five hypotheses were formulated with quantitative targets and tested against a live being with 66,163 knowledge triples. None of the five hypotheses met their targets. The HDD methodology delivered its core value: fast identification of the root cause within weeks rather than months, enabling rapid redirect of engineering effort.