What is TAME in Biology?

PRINT ENGLISH BIOELECTRICITY GUIDE

PRINT CHINESE BIOELECTRICITY GUIDE


What is TAME in Biology? Summary

  • TAME = Technological Approach to Mind Everywhere: It’s a framework, developed by Michael Levin and collaborators, for understanding and interacting with *diverse forms of intelligence*, regardless of their physical basis (biological, artificial, etc.).
  • Beyond Human-Centric Thinking: TAME challenges the assumption that human-like intelligence is the only “true” form of intelligence. It embraces a broader, more inclusive view.
  • A Gradualist View of Cognition: TAME doesn’t see a sharp line between “cognitive” and “non-cognitive” systems. Instead, it proposes a *spectrum* of cognitive abilities, from simple to complex.
  • The “Axis of Persuadability”: A key concept is the *axis of persuadability* – how easily a system’s behavior can be changed, ranging from physical rewiring (like a clock) to rational argument (like a human).
  • Focus on Goals: TAME emphasizes *goal-directed behavior* as a fundamental characteristic of intelligent systems, even at very basic levels (like cells).
  • “Selves”. From single molecule to entire bodies, all exhibit different goal pursuits at their level of organization; selves can be, essentially, assigned according to competency at some relevant “goal” or “agenda”. This offers powerful implications for AI, for morphogenesis.
  • The “Cognitive Light Cone”: This concept describes the scale of a system’s goals and its ability to influence its environment. A single cell has a small light cone; a multicellular organism has a larger one.
  • Morphogenesis as Cognition: TAME views the development and regeneration of body form (morphogenesis) as a form of *basal cognition*, where cell collectives exhibit goal-directed behavior in “morphospace.”
  • Bioelectricity’s Role: Bioelectric signaling, particularly through gap junctions, is seen as a key mechanism for *scaling up* cognition, allowing individual cells to work together towards larger-scale goals.
  • Practical Applications: TAME has implications for regenerative medicine (controlling tissue growth), artificial intelligence (designing more adaptable systems), and even ethics (reconsidering our moral obligations to different forms of intelligence).

Beyond “Us vs. Them”: A New Framework for Intelligence

We often have a very narrow view of intelligence, limiting it to humans or animals with complex brains. We tend to think of a sharp dividing line between “intelligent” beings (like us) and “non-intelligent” things (like rocks, plants, or simple machines). But what if this is the wrong way to think about it?

TAME – the Technological Approach to Mind Everywhere – is a framework developed by Michael Levin and his collaborators (notably, the philosopher Thomas Metzinger was also involved) that challenges this narrow view. TAME proposes a much broader, more inclusive, and more *practical* way of understanding and interacting with intelligence in *all* its forms, whether it’s biological, artificial, or something else entirely.


A Gradualist View: The Spectrum of Cognition

One of the core ideas of TAME is *gradualism*. Instead of a sharp line between “intelligent” and “non-intelligent” systems, TAME proposes a *spectrum* of cognitive abilities. This relates closely with concepts such as Basal Cognition, and scaling up/scaling down of cellular intelligence.

At one end of the spectrum, you have very simple systems, like a thermostat, that have very limited ability to process information or adapt their behavior. At the other end, you have complex systems, like human brains, that can reason, plan, learn, and adapt in incredibly sophisticated ways.

But *in between*, there’s a vast range of systems with varying degrees of cognitive ability: bacteria, plants, slime molds, insect colonies, animal brains of different sizes and complexities, and even *cells and tissues* within our own bodies.

The levels of organization:

  • Molecular
  • Pathways
  • Cells, tissues, whole body: exhibiting cognitive abilities on different, increasing “cognitive light cone” scope sizes.
  • Bioelectricity has very strong effects on “integrating”, or binding the units below, into unified action/cognition capabilities.

The “Axis of Persuadability”: How to Influence a System

A key concept in TAME is the *axis of persuadability*. This describes how easy or difficult it is to change a system’s behavior, and *what kind* of intervention is required to do so.

Consider these examples:

  • A Rock: To change the shape of a rock, you need to apply brute physical force – chipping, grinding, etc. It’s at the very low end of the persuadability axis.
  • A Clock: To change the time displayed by a clock, you need to physically manipulate its mechanism – move the hands or adjust the settings. It’s slightly more persuadable than a rock, but still requires physical intervention.
  • A Simple Robot: You can reprogram a simple robot by changing its software code. It’s more persuadable than a clock, as you don’t need to physically alter its hardware.
  • An Animal: You can train an animal using rewards and punishments. It’s more persuadable than a robot, as its behavior can be shaped by experience.
  • A Human: You can persuade a human using rational argument, evidence, and emotional appeals. This is the highest end of the persuadability axis.
  • The degree by which a system can be influenced or altered defines how much “agency”.
  • For simple hardware: only possible physical intervention is breaking/rebuilding.
  • For more complex tissues, organs: training and new experiences alter its ability to process future inputs.

TAME suggests that the position of a system on this axis is a good indicator of its level of cognitive sophistication. The *more* ways a system can be influenced, and the *more abstract* those ways are (moving from physical force to information and communication), the more “intelligent” the system is.


Goal-Directed Behavior: The Hallmark of “Mind”

Another central idea in TAME is that *goal-directed behavior* is a fundamental characteristic of intelligent systems, even at very basic levels.

A “goal” doesn’t have to be a complex, conscious desire. It can be as simple as a bacterium moving towards a source of nutrients or a cell migrating to a wound site to repair damage. These examples involve stress signals, memory, and even learning – that’s where the term, cognition can reasonably applied.

TAME proposes that any system that exhibits goal-directed behavior – that acts in a way that tends to achieve a particular outcome – can be considered to have a “mind,” at least to some degree. This is a very broad definition of “mind,” but it’s a useful one for understanding the continuum of cognitive abilities in the natural world.

This involves not just individual cells, but also a collection (and “collections of collections”). For example, a group of gap-junction cells connected into one electrical network exhibit new computational properties. They may start “caring about” states, problems or other issues far outside just their narrow, previous scope; this emergent, bigger unit thus gains larger intelligence.


The “Cognitive Light Cone”: Scaling Up Goals

The *cognitive light cone* concept, which we’ve discussed before, is closely related to goal-directedness. It describes the scale of a system’s goals and its ability to influence its environment.

  • A *single cell* has a relatively small cognitive light cone. Its goals are primarily focused on its own survival and immediate surroundings.
  • A *multicellular organism* has a much larger cognitive light cone. Its goals can include things like finding food, avoiding predators, reproducing, and even building complex structures (like nests or hives).

Bioelectricity, particularly through gap junctions, plays a crucial role in *expanding* the cognitive light cone. By allowing cells to share information and coordinate their actions, gap junctions effectively create a larger, more intelligent “self.”


Morphogenesis as Cognition: Building Bodies with “Thought”

One of the most radical implications of TAME is that we can view *morphogenesis* – the development and regeneration of body form – as a form of *basal cognition*.

Even without neural-level pathways, cells and tissues can also process stress signals, memories. A frog, after cutting, rebuild itself toward intended form. When face structures are arranged, a developing froglet corrects errors. These surprising feats are very “goal-directed” – and that is why Michael Levin uses “cognitive-like” explaination model for them. It is a form of intelligence without any obvious centralized controllers/”brain”.

During development, cells collectively “solve the problem” of building a complex, functional organism. They do this by sensing their environment, communicating with each other, and adjusting their behavior to achieve a “target morphology” – the desired shape and structure of the body.

This is not just a pre-programmed, mechanical process. It’s a dynamic, adaptive, *goal-directed* process, where cells exhibit a kind of “intelligence” in their ability to coordinate their actions and achieve a specific outcome.


Practical Implications of TAME

TAME is not just a philosophical framework; it has practical implications for many fields:

  • Regenerative Medicine: By understanding morphogenesis as a cognitive process, we can develop new strategies for controlling tissue growth and regeneration. We might be able to “persuade” cells to rebuild lost limbs or organs by providing the right bioelectric “instructions.”
  • Artificial Intelligence: TAME suggests that we can build more adaptable and intelligent AI systems by mimicking the principles of biological cognition, such as goal-directedness, distributed information processing, and collective intelligence.
  • Ethics: TAME challenges us to reconsider our moral obligations to different forms of intelligence. If even simple biological systems exhibit some degree of “mind,” how should we treat them? This is particularly relevant as we develop increasingly sophisticated AI systems and bioengineered constructs.
  • Cancer: From being cells going into a rogue group to revert to ancient ways of single cell survival (by unlinking themselves from neighboring gap junction networks). This could involve how cancers ignore those normal “instructions”.
  • Xenobots: These frog skin cells (not genetically modified, with typical frog cell components) can spontaneously create “creatures”, without any brains or typical neurons. How do cells accomplish such complex re-arrangement when cut from tissue? The latent capability found from such biological system demonstrates very clearly how cognitive behaviours at level of tissue – not merely just as single-celled microbes – can work toward a larger “outcome”.
  • Anthromorphs: Studies showing, like xenobots, when given different pathways (being freed from their normal, default developmental process) can develop behaviours far from original tissues purpose. These “human cells” were also found demonstrating complex problem-solving traits.

TAME offers a powerful new way of thinking about intelligence, not as a unique human property, but as a fundamental aspect of life, present in diverse forms and at multiple scales. It opens up exciting new avenues for research and has the potential to transform many fields, from medicine to technology to our understanding of ourselves.


什么是生物学中的 TAME?摘要

  • TAME = 随处可见心智的技术方法 (Technological Approach to Mind Everywhere):这是一个框架,由 Michael Levin 及其合作者开发,用于理解和与*不同形式的智能*互动,无论其物理基础是什么(生物、人工等)。
  • 超越以人为中心的思维: TAME 挑战了类似人类的智能是唯一“真正”智能形式的假设。它拥抱更广泛、更具包容性的观点。
  • 认知的渐进观: TAME 并没有看到“认知”和“非认知”系统之间存在明显的界限。相反,它提出了一个从简单到复杂的*认知能力谱系*。
  • “可说服性轴”: 一个关键概念是*可说服性轴*—— 改变系统行为的难易程度,从物理重新布线(如时钟)到理性论证(如人类)。
  • 关注目标: TAME 强调*目标导向行为*是智能系统的基本特征,即使在非常基本的水平(如细胞)也是如此。
  • “自我”。从单个分子到整个身体,所有这些都在其组织水平上表现出不同的目标追求;本质上,可以根据某些相关“目标”或“议程”的能力来分配自我。这为人工智能、形态发生提供了强大的启示。
  • “认知光锥”: 这个概念描述了系统目标的规模及其影响环境的能力。单个细胞具有较小的光锥;多细胞生物具有较大的光锥。
  • 形态发生作为认知: TAME 将身体形态的发育和再生(形态发生)视为一种*基础认知*形式,其中细胞集体在“形态空间”中表现出目标导向行为。
  • 生物电的作用: 生物电信号,特别是通过间隙连接,被认为是*扩大*认知的关键机制,允许单个细胞协同工作以实现更大规模的目标。
  • 实际应用: TAME 对再生医学(控制组织生长)、人工智能(设计更具适应性的系统),甚至伦理学(重新考虑我们对不同形式智能的道德义务)都有影响。

超越“我们 vs. 他们”:一个新的智能框架

我们通常对智能有一个非常狭隘的看法,将其限制在人类或具有复杂大脑的动物身上。我们倾向于认为“智能”生物(如我们)和“非智能”事物(如岩石、植物或简单的机器)之间存在明显的分界线。但如果这是一种错误的思考方式呢?

TAME —— 随处可见心智的技术方法 —— 是由 Michael Levin 及其合作者(值得注意的是,哲学家 Thomas Metzinger 也参与其中)开发的一个框架,挑战了这种狭隘的观点。TAME 提出了一种更广泛、更具包容性、更*实用*的方式来理解和与*所有*形式的智能互动,无论是生物的、人工的,还是完全不同的东西。


渐进观:认知谱系

TAME 的核心思想之一是*渐进主义*。TAME 并没有在“智能”和“非智能”系统之间划清界限,而是提出了一个*认知能力谱系*。这与基础认知以及细胞智能的放大/缩小等概念密切相关。

在这个谱系的一端,你有非常简单的系统,比如恒温器,它们处理信息或调整其行为的能力非常有限。在另一端,你有复杂的系统,比如人脑,它们可以以极其复杂的方式推理、计划、学习和适应。

但在*两者之间*,存在着各种各样的系统,它们具有不同程度的认知能力:细菌、植物、粘菌、昆虫群落、不同大小和复杂性的动物大脑,甚至我们自己体内的*细胞和组织*。

组织层次:

  • 分子
  • 通路
  • 细胞、组织、整个身体:在不同的、不断增加的“认知光锥”范围大小上表现出认知能力。
  • 生物电对“整合”或结合下面的单元成为统一的行动/认知能力具有非常强大的影响。

“可说服性轴”:如何影响一个系统

TAME 中的一个关键概念是*可说服性轴*。这描述了改变系统行为的难易程度,以及需要进行*哪种*干预才能做到这一点。

考虑以下示例:

  • 一块岩石: 要改变岩石的形状,你需要施加蛮力 —— 削、磨等。它处于可说服性轴的最低端。
  • 一个时钟: 要更改时钟显示的时间,你需要物理操纵其机械装置 —— 移动指针或调整设置。它比岩石稍微更具说服性,但仍然需要物理干预。
  • 一个简单的机器人: 你可以通过更改其软件代码来重新编程一个简单的机器人。它比时钟更具说服性,因为你不需要物理改变它的硬件。
  • 一只动物: 你可以使用奖励和惩罚来训练动物。它比机器人更具说服性,因为它的行为可以通过经验来塑造。
  • 一个人: 你可以使用理性论证、证据和情感诉求来说服一个人。这是可说服性轴的最高端。
  • 系统可以被影响或改变的程度定义了多少“能动性”。
  • 对于简单的硬件:只有可能的物理干预是破坏/重建。
  • 对于更复杂的组织、器官:训练和新的经验会改变其处理未来输入的能力。

TAME 认为,一个系统在这个轴上的位置很好地表明了它的认知复杂程度。一个系统可以被影响的方式越*多*,这些方式越*抽象*(从物理力到信息和沟通),系统就越“智能”。


目标导向行为:“心智”的标志

TAME 的另一个核心思想是,*目标导向行为*是智能系统的基本特征,即使在非常基本的水平也是如此。

“目标”不一定是复杂的、有意识的愿望。它可以像细菌朝着营养源移动或细胞迁移到伤口部位以修复损伤一样简单。这些例子涉及压力信号、记忆,甚至学习 —— 这就是术语“认知”可以合理应用的地方。

TAME 提出,任何表现出目标导向行为的系统 —— 其行为趋向于实现特定结果 —— 都可以被认为具有某种程度的“心智”。这是一个非常广泛的“心智”定义,但它对于理解自然界中认知能力的连续性非常有用。

这不仅涉及单个细胞,还涉及集合(和“集合的集合”)。例如,一组连接到一个电网络的间隙连接细胞表现出新的计算特性。它们可能开始“关心”状态、问题或其他远远超出其狭窄的先前范围的问题;这种涌现的、更大的单元从而获得了更大的智能。


“认知光锥”:扩大目标

我们之前讨论过的*认知光锥*概念与目标导向性密切相关。它描述了系统目标的规模及其影响环境的能力。

  • *单个细胞*具有相对较小的认知光锥。它的目标主要集中在自身的生存和周围环境。
  • *多细胞生物*具有大得多的认知光锥。它的目标可以包括寻找食物、躲避捕食者、繁殖,甚至建造复杂的结构(如巢穴或蜂巢)。

生物电,特别是通过间隙连接,在*扩大*认知光锥方面起着至关重要的作用。通过允许细胞共享信息和协调它们的行动,间隙连接有效地创造了一个更大、更智能的“自我”。


形态发生作为认知:用“思想”构建身体

TAME 最激进的含义之一是,我们可以将*形态发生* —— 身体形态的发育和再生 —— 视为一种*基础认知*形式。

即使没有神经水平的通路,细胞和组织也可以处理压力信号、记忆。青蛙在切割后会重建自身以达到预期的形态。当面部结构被排列时,发育中的蝌蚪会纠正错误。这些令人惊讶的壮举非常“目标导向”—— 这就是为什么 Michael Levin 使用“类认知”解释模型来解释它们。这是一种没有任何明显的中枢控制器/“大脑”的智能形式。

在发育过程中,细胞集体“解决”构建复杂、功能性生物体的问题。它们通过感知环境、相互沟通并调整其行为以实现“目标形态”—— 身体所需的形状和结构来做到这一点。

这不仅仅是一个预先编程的机械过程。这是一个动态的、适应性的、*目标导向*的过程,细胞在其协调行动和实现特定结果的能力中表现出一种“智能”。


TAME 的实际意义

TAME 不仅仅是一个哲学框架;它对许多领域都有实际意义:

  • 再生医学: 通过将形态发生理解为一个认知过程,我们可以开发控制组织生长和再生的新策略。我们或许能够通过提供正确的生物电“指令”来“说服”细胞重建失去的四肢或器官。
  • 人工智能: TAME 表明,我们可以通过模仿生物认知的原理(如目标导向性、分布式信息处理和集体智慧)来构建更具适应性和智能的 AI 系统。
  • 伦理学: TAME 挑战我们重新考虑我们对不同形式智能的道德义务。如果即使是简单的生物系统也表现出一定程度的“心智”,我们应该如何对待它们?随着我们开发出越来越复杂的人工智能系统和生物工程结构,这一点尤其重要。
  • 癌症: 从细胞进入流氓群体恢复到古代的单细胞生存方式(通过将自己与相邻的间隙连接网络断开)。这可能涉及癌症如何忽略那些正常的“指令”。
  • 异种机器人: 这些青蛙皮肤细胞(未进行基因改造,具有典型的青蛙细胞成分)可以自发地创造“生物”,而无需任何大脑或典型的神经元。当从组织中切割出来时,细胞如何完成如此复杂的重新排列?在这种生物系统中发现的潜在能力非常清楚地表明,组织水平的认知行为 —— 不仅仅是作为单细胞微生物 —— 如何朝着更大的“结果”努力。
  • Anthromorphs:研究表明,与异种机器人一样,当给予不同的途径时(从其正常的、默认的发育过程中解放出来),可以发展出远离原始组织目的的行为。这些“人类细胞”也被发现表现出复杂的问题解决特征。

TAME 提供了一种思考智能的强大新方式,不是将其视为人类独有的属性,而是将其视为生命的一个基本方面,以不同的形式和在多个尺度上存在。它为研究开辟了令人兴奋的新途径,并有可能改变许多领域,从医学到技术,再到我们对自身的理解。