本文探讨了将大脑视为心智的益处和局限性。心智模型简化了我们对人类行为的理解，心理学和哲学也广泛使用它。然而，心智模型也可能导致误判，例如期望人们拥有始终如一的观点。为了避免这些陷阱，作者提倡一种双重模型，即同时使用心智和大脑两种模型来理解人类行为。作者用“心智是魔法”来提醒自己，不要忽视与心智模型不符的现象，允许自己在使用心智模型和大脑模型之间切换。在实践中，作者主要使用心智模型，但会留出空间来容纳不符合该模型的现象。

🧠 大脑是神经元的复杂连接，虽然物理学角度是确定的，但我们更倾向于将其视为“心智”，一个整洁且一致的模型。这种视角在心理学和哲学中被广泛应用，有助于我们更好地理解和预测人类行为。

💡 将大脑视为心智有很多好处，例如，研究表明，拥有内在控制点和成长型思维模式的人更容易成功。从个人经验来看，当把人看作心智时，更容易预测他们的行为，并且有些预测非常接近现实。

⚠️ 然而，将大脑视为心智也存在缺陷，可能导致一系列误判。例如，期望人们持有始终如一的观点，而忽视了他们可能拥有一系列松散连接的态度。此外，还可能出现盲点偏差，以及期望一次性学习就能掌握所有知识。

✨ 为了避免心智模型的误判，作者提出了一种双重模型。这种模型允许我们同时使用心智和大脑两种模型来理解人类行为。作者用“心智是魔法”来提醒自己，不要忽视那些与心智模型不符的现象。

Published on May 16, 2025 7:10 PM GMT

As we know, brains are incredibly complex connections of neurons.  If physics is deterministic so are our brains.  But that's not how we think about brains.  We think about them as MINDS.  And minds are neat and consistent.  

There are a lot of benefits to thinking about brains as minds:

Studies have shown higher success for people with internal loci of control and for individuals with growth mindsets.  From personal experience, I have a much easier time making predictions about people when I think of them as minds, and some of those predictions even come close to reality.Much of psychology and philosophy make use of the MIND model.
 

However, there are also drawbacks to thinking about brains as minds.  Here is a very non-exhaustive list of mispredictions that can happen:

Expecting people to have consistent views instead of a collection of loosely connected attitudes."He loves me.  He loves me not." instead of "He posseses some neural circuits I would classify as 'loves me' which activate reliably upon thinking about me."The bias blindspot.Expecting to only need to learn something once instead of "maybe in my current state, this information stimulus will create beneficial neural connection changes".Expecting short inferential distances.

Note: Each of those points is potentially very profound and useful (as in each has been a big aha moment for me at some point), but most likely inert to any given reader at any given time (see #4).

 

So, MINDS are a useful model, but don't match reality in some important ways.  How can we use the model of minds while avoiding the pitfalls of misprediction?
One way to do this is to carve out space for a dual-model (or ensemble model if you prefer 3+).  We exhibit mind-brain duality (and like wave-particle duality, probably neither model matches a true technical understanding).  

One phrase that helps me with this, is "Minds are magic".  Insert your own phrase that does it for you.  For me, it means "don't discard the phenomena I've observed that don't fit with the MIND model.  Instead, allow yourself the double-think of conceptualizing people as minds when that works and brains when that works better."  It's also a reminder that I face a halting-problem-like barrier to being able to truly model a brain inside my brain, and that it's okay that "the map is not the territory" because "it fits inside your glove compartment".

In practice - I use the MIND model >99% of the time, so I have to carve out a mental space for allowing the phenomena that don't fit with the MIND model.  If you use a different model, you'll need a different carve out.  
"Minds are magic" disarms my sense of "this should all seem 100% consistent to me". And the fact that it works for me, is Magic.

Discuss