![\"AI](\"https:\/\/www.newsbeep.com\/ca\/wp-content\/uploads\/2026\/01\/ai-novel-biological-organism.jpg\")
<\/p>\n\t<\/p>\n
\t\t\tIllustration by Tag Hartman-Simkins \/ Futurism. Source: Getty Images<\/p>\n
AI models are now everywhere, from hospitals<\/a> to churches<\/a>.<\/p>\n The astonishing thing is that even AI experts still don\u2019t know exactly what\u2019s happening inside these black box<\/a> models, even as they\u2019re being deployed in the highest-stakes settings imaginable<\/a>.The latest strategy to figure it out: studying them like biological systems.<\/p>\n For example, MIT Tech Review reports<\/a>, scientists at Anthropic have developed tools that let them trace what\u2019s happening inside models as they perform a task, a type of study called mechanistic interpretability \u2014 which resembles how doctors use MRIs<\/a> to study brain activity, another type of intelligence we don\u2019t quite understand yet.<\/p>\n \u201cThis is very much a biological type of analysis,\u201d Josh Batson, a research scientist at Anthropic, told Tech Review. \u201cIt\u2019s not like math or physics.\u201d<\/p>\n In another experiment that resembles how biologists use organoids<\/a>, which are miniature versions of human organs, the magazine reports that Anthropic developed a special neural network called a sparse autoencoder whose inner workings are easier to understand and analyze than regular large language models (LLMs).<\/p>\n Another technique is chain-of-thought monitoring, in which models explain their reasoning behind their behavior and actions \u2014 much like listening to the inner monologue of an actual person. This has helped scientists spot misaligned behavior.<\/p>\n \u201cIt\u2019s been pretty wildly successful in terms of actually being able to find the model doing bad things,\u201d said Bowen Baker, OpenAI research scientist, to MIT.<\/p>\n A looming danger is that future models will become so complex \u2014 especially if they\u2019re themselves designed by AI \u2014 that we\u2019ll have effectively no idea how they work. Even now, with the current tools and techniques at our disposal, unexpected behaviors still pop up that don\u2019t align with human objectives of truthfulness and safety.<\/p>\n We see hard evidence of this in the news, which is littered with reports of people harming themselves<\/a> because AI told them to \u2014 which makes it even more disturbing that we still don\u2019t quite understand how they function.<\/p>\n