Lean Hypothesis Testing in Practice: Five Rapid Experiments with Predictive Processing for Neurosymbolic AI
Fuente:
Zenodo
Guardado en:
| Autor principal: | |
|---|---|
| Formato: | Recurso digital |
| Publicado: |
Zenodo
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866901818340016128 |
|---|---|
| author | Head, Hank |
| author_facet | Head, Hank |
| contents | 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. |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_19788392 |
| institution | Zenodo |
| language | |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Lean Hypothesis Testing in Practice: Five Rapid Experiments with Predictive Processing for Neurosymbolic AI Head, Hank hypothesis-driven development predictive processing neurosymbolic AI lean experimentation negative results 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. |
| title | Lean Hypothesis Testing in Practice: Five Rapid Experiments with Predictive Processing for Neurosymbolic AI |
| topic | hypothesis-driven development predictive processing neurosymbolic AI lean experimentation negative results |
| url | https://doi.org/10.5281/zenodo.19788392 |