# Learning-like behavior deserves precise language

Proofs of concept. DishBrain coupled cultured human or rodent neurons in vitro to a simulated Pong environment in silico through a multielectrode interface. It is an important example of closed-loop interaction between living neural cultures and software. These were neuronal cultures, not a restored human brain. The paper’s use of the word sentience should not be presented as settled proof of subjective experience or human-level understanding.

Canonical page: https://givegigs.com/wiki/necro-chapter-biocomputing
Last updated: 2026-11-09T18:58:11.921Z
Parent topic: [necromancer-neuroscience](https://givegigs.com/wiki/necromancer-neuroscience)

## Contents
- [Overview](#wiki-section-overview-1)
- [Chapter notes](#wiki-section-chapter-notes-2)
- [Dossier section](#wiki-section-dossier-link-3)

<a id="wiki-section-overview-1"></a>
## Overview

Proofs of concept. This page expands the “Learning-like behavior deserves precise language” chapter in the neuroscience collection.

<a id="wiki-section-chapter-notes-2"></a>
## Chapter notes

DishBrain coupled cultured human or rodent neurons in vitro to a simulated Pong environment in silico through a multielectrode interface. It is an important example of closed-loop interaction between living neural cultures and software. These were neuronal cultures, not a restored human brain. The paper’s use of the word sentience should not be presented as settled proof of subjective experience or human-level understanding.

Brainoware, reported in 2023, used a brain organoid as a biological reservoir in a computing system and evaluated tasks including speech recognition and nonlinear prediction. The result concerns a particular hybrid experimental architecture. It does not establish general intelligence, a transferable persona, indefinite survival, or a verified advantage for every computing workload.

A 2024 neural-organoid study measured features relevant to basic learning and memory, including plasticity-associated gene expression and network dynamics. Such work helps turn broad concepts into biological measurements. The startup’s proposed standard should be to describe the task, baseline, training history, controls, and uncertainty before adopting terms such as learning, memoria, or intelligence.

<a id="wiki-section-dossier-link-3"></a>
## Dossier section

[Open this exact section in the Necromancer dossier](/necromancer#biocomputing). The original chapter status is “Proofs of concept.”

## Sources

- [In vitro neurons learn and exhibit sentience when embodied in a simulated game-world](https://pubmed.ncbi.nlm.nih.gov/36228614/)
- [Brain organoid reservoir computing for artificial intelligence](https://www.nature.com/articles/s41928-023-01069-w)
- [Human Neural Organoid Microphysiological Systems Show the Building Blocks Necessary for Basic Learning and Memory](https://pubmed.ncbi.nlm.nih.gov/39345518/)

---
This article is collaboratively maintained. Check the canonical page and cited sources for the latest context.