Center for Alternatives to Animal Testing (CAAT), Johns Hopkins University, Bloomberg School of Public Health, Baltimore, MD, USA.
CAAT Europe, University of Konstanz, Konstanz, Germany.
ALTEX. 2023;40(2):191-203. doi: 10.14573/altex.2303261.
Understanding brain function remains challenging as work with human and animal models is complicated by compensatory mechanisms, while in vitro models have been too simple until now. With the advent of human stem cells and the bioengineering of brain microphysiological systems (MPS), understanding how both cognition and long-term memory arise is now coming into reach. We suggest combining cutting-edge AI with MPS research to spearhead organoid intelligence (OI) as synthetic biological intelligence. The vision is to realize cognitive functions in brain MPS and scale them to achieve relevant short- and long-term memory capabilities and basic information processing as the ultimate functional experimental models for neurodevelopment and neurological function and as cell-based assays for drug and chemical testing. By advancing the frontiers of biological computing, we aim to (a) create models of intelligence-in-a-dish to study the basis of human cognitive functions, (b) provide models to advance the search for toxicants contributing to neurological diseases and identify remedies for neurological maladies, and (c) achieve relevant biological computational capacities to complement traditional computing. Increased understanding of brain functionality, in some respects still superior to today's supercomputers, may allow to imitate this in neuromorphic computer architectures or might even open up biological computing to complement silicon computers. At the same time, this raises ethical questions such as where sentience and consciousness start and what the relationship between a stem cell donor and the respective OI system is. Such ethical discussions will be critical for the socially acceptable advance of brain organoid models of cognition.
理解大脑功能仍然具有挑战性，因为人类和动物模型的研究受到代偿机制的影响，而体外模型直到现在还过于简单。随着人类干细胞的出现和脑微生理系统（MPS）的生物工程化，理解认知和长期记忆是如何产生的现在已经成为可能。我们建议将最先进的人工智能与 MPS 研究相结合，引领类器官智能（OI）作为合成生物学智能。我们的目标是在脑 MPS 中实现认知功能，并对其进行扩展，以实现相关的短期和长期记忆能力以及基本信息处理，作为神经发育和神经功能的最终功能实验模型，以及作为基于细胞的药物和化学测试的筛选模型。通过推进生物计算的前沿，我们旨在（a）创建智能器官模型，以研究人类认知功能的基础，（b）提供模型以推进对导致神经疾病的毒物的研究，并确定治疗神经疾病的方法，以及（c）实现相关的生物计算能力，以补充传统计算。对大脑功能的进一步理解，在某些方面仍然优于当今的超级计算机，这可能允许在神经形态计算机架构中模拟这一点，甚至可能为生物计算机打开补充硅计算机的途径。与此同时，这引发了一些伦理问题，例如意识和意识从哪里开始，以及干细胞供体与相应的 OI 系统之间的关系是什么。这些伦理讨论对于可接受的社会进步的大脑类器官认知模型至关重要。
