The amazing brains and morphing skin of octopuses and other cephalopods Roger Hanlon

This is a strange and wonderful brain,

one that gives rise to an idea
of a kind of alternative intelligence

on this planet.

This is a brain that is formed
in a very strange body,

one that has the equivalent
of small satellite brains

distributed throughout that body.

How different is it from the human brain?

Very different, so it seems,

so much so that my colleagues and I
are struggling to understand

how that brain works.

But what I can tell you for certain

is that this brain is capable
of some amazing things.

So, who does this brain belong to?

Well, join me for a little bit
of diving into the ocean,

where life began,

and let’s have a look.

You may have seen some of this before,

but we’re behind a coral reef,
and there’s this rock out there,

a lot of sand, fishes swimming around …

And all of a sudden this octopus appears,

and now it flashes white,

inks in my face and jets away.

In slow motion reverse,

you see the ring develop around the eye,

and then the pattern develops in the skin.

And now watch the 3-D texture
of the skin change

to really create this
beautiful, 3-D camouflage.

So there are 25 million color organs
called “chromatophores” in the skin,

and all those bumps out there,
which we call “papillae,”

and they’re all neurally controlled
and can change instantaneously.

I would argue that dynamic camouflage

is a form of “intelligence.”

The level of complexity of the skin
with fast precision change

is really quite astonishing.

So what can you do with this skin?

Well, let’s think a little bit
about other things besides camouflage

that they can do with their skin.

Here you see the mimic octopus
and a pattern.

All of a sudden,
it changes dramatically –

that’s signaling, not camouflage.

And then it goes back
to the normal pattern.

Then you see the broadclub cuttlefish

showing this passing cloud display
as it approaches a crab prey.

And finally, you see the flamboyant
cuttlefish in camouflage

and it can shift instantly
to this bright warning display.

What we have here
is a sliding scale of expression,

a continuum, if you will,

between conspicuousness and camouflage.

And this requires a lot of control.

Well, guess what?

Brains are really good for control.

The brain of the octopus shown here
has 35 lobes to the brain,

80 million tiny cells.

And even though that’s interesting,

what’s really odd is that
the skin of this animal

has many more neurons, as illustrated
here, especially in the yellow.

There are 300 million neurons in the skin

and only 80 million in the brain itself –

four times as many.

Now, if you look at that,

there’s actually one of those
little satellite brains

and the equivalent of the spinal cord
for each of the eight arms.

This is a very unusual way
to construct a nervous system in a body.

Well, what is that brain good for?

That brain has to outwit
other big, smart brains

that are trying to eat it,

and that includes porpoises and seals

and barracudas and sharks

and even us humans.

So decision-making is one of the things
that this brain has to do,

and it does a very good job of it.

Shown here, you see this octopus
perambulating along,

and then it suddenly stops
and creates that perfect camouflage.

And it’s really marvelous,

because when these animals
forage in the wild,

they have to make over a hundred
camouflaging decisions

in a two-hour forage,

and they do that twice a day.

So, decision-making.

They’re also figuring out where to go
and how to get back home.

So it’s a decision-making thing.

We can test this camouflage,

like that cuttlefish you see behind me,

where we pull the rug out from under it

and give it a checkerboard,

and it even uses that strange
visual information

and does its best to match the pattern
with a little ad-libbing.

So other cognitive skills
are important, too.

The squids have a different
kind of smarts, if you will.

They have an extremely complex,
interesting sex life.

They have fighting and flirting
and courting and mate-guarding

and deception.

Sound familiar?

(Laughter)

And it’s really quite amazing

that these animals have
this kind of intuitive ability

to do these behaviors.

Here you see a male and a female.

The male, on the left,
has been fighting off other males

to pair with the female,

and now he’s showing a dual pattern.

He shows courtship and love on her side,

fighting on the other.

Watch him when she shifts places –

(Laughter)

and you see that he has fluidly
changed the love-courtship pattern

to the side of the female.

So this kind of dual signaling
simultaneously

with a changing behavioral context

is really extraordinary.

It takes a lot of brain power.

Now, another way
to look at this is that, hmm,

maybe we have 50 million years
of evidence for the two-faced male.

(Laughter)

All right, let’s move on.

(Laughter)

An octopus on a coral reef
has a tough job in front it

to go to so many places,
remember and find its den.

And they do this extremely well.

They have short- and long-term memory,

they learn things
in three to five trials –

it’s a good brain.

And the spatial memory is unusually good.

They will even end their forage
and make a beeline

all the way back to their den.

The divers watching them
are completely lost,

but they can get back,

so it’s really quite refined
memory capability.

Now, in terms of cognitive skills,

look at this sleeping behavior
in the cuttlefish.

Especially on the right,
you see the eye twitching.

This is rapid eye movement
kind of dreaming

that we only thought
mammals and birds did.

And you see the false color
we put in there

to see the skin patterning flashing,

and this is what’s happening a lot.

But it’s not normal awake behaviors;
it’s all different.

Well, dreaming is when you have
memory consolidation,

and so this is probably
what’s happening in the cuttlefish.

Now, another form of memory
that’s really unusual

is episodic-like memory.

This is something that humans need
four years of brain development to do

to remember what happened
during a particular event,

where it happened and when it happened.

The “when” part is particularly difficult,

and these children can do that.

But guess what?

We find recently that the wily cuttlefish
also has this ability,

and in experiments last summer,

when you present a cuttlefish
with different foods at different times,

they have to match that
with where it was exactly

and when was the last time they saw it.

Then they have to guide their foraging
to the rate of replenishment

of each food type in a different place.

Sound complicated?

It’s so complicated, I hardly
understood the experiment.

So this is really high-level
cognitive processing.

Now, speaking of brains
and evolution at the moment,

you look on the right, there’s the pathway
of vertebrate brain evolution,

and we all have good brains.

I think everyone will acknowledge that.

But if you look on the left side,

some of the evolutionary pathway
outlined here to the octopus,

they have both converged, if you will,
to complex behaviors

and some form of intelligence.

The last common denominator
in these two lines

was 600 million years ago,

and it was a worm with very few neurons,

so very divergent paths

but convergence of complicated behavior.

Here is the fundamental question:

Is the brain structure of an octopus

basically different
down to the tiniest level

from the vertebrate line?

Now, we don’t know the answer,

but if it turns out to be yes,

then we have a different
evolutionary pathway

to create intelligence on planet Earth,

and one might think that
the artificial intelligence community

might be interested in those mechanisms.

Well, let’s talk genetics
just for a moment.

We have genomes, we have DNA,

DNA is transcripted into RNA,

RNA translates that into a protein,
and that’s how we come to be.

Well, the cephalopods do it differently.

They have big genomes, they have DNA,

they transcript it into RNA,

but now something
dramatically different happens.

They edit that RNA
at an astronomical weird rate,

a hundredfold more than
we as humans or other animals do.

And it produces scores of proteins.

And guess where most of them are for?

The nervous system.

So perhaps this is an unorthodox way

for an animal to evolve
behavioral plasticity.

This is a lot of conjecture,
but it’s food for thought.

Now, I’d like to share
with you for a moment

my experience, and using my smarts
and that of my colleagues,

to try and get this kind of information.

We’re diving, we can’t stay
underwater forever

because we can’t breathe it,

so we have to be efficient in what we do.

The total sensory immersion
into that world

is what helps us understand
what these animals are really doing,

and I have to tell you that
it’s really an amazing experience

to be down there and having
this communication

with an octopus and a diver

when you really begin to understand
that this is a thinking, cogitating,

curious animal.

And this is the kind of thing
that really inspires me endlessly.

Let’s go back to that smart skin
for a few moments.

Here’s a squid and a camouflage pattern.

We zoom down and we see
there’s beautiful pigments and reflectors.

There are the chromatophores
opening and closing very quickly.

And then, in the next layer of skin,

it’s quite interesting.

The chromatophores are closed,

and you see this magical iridescence
just come out of the skin.

This is also neurally controlled,

so it’s the combination of the two,

as seen here in the high-resolution
skin of the cuttlefish,

where you get this beautiful
pigmentary structural coloration

and even the faint blushing
that is so beautiful.

Well, how can we make use
of some of this information?

I talked about those
skin bumps, the papillae.

Here’s the giant Australian cuttlefish.

It’s got smooth skin
and a conspicuous pattern.

I took five pictures in a row
one second apart,

and just watch this animal morph –
one, two, three, four, five –

and now I’m a seaweed.

And then we can come right back out of it

to see the smooth skin
and the conspicuousness.

So this is really
marvelous, morphing skin.

You can see it in more detail here.

Periscope up,

and you’ve got those beautiful papillae.

And then we look in a little more detail,

you can see the individual
papillae come up,

and there are little ridges on there,

so it’s a papilla on papilla and so forth.

Every individual species out there
has more than a dozen shapes and sizes

of those bumps

to create fine-tuned,
neurally controlled camouflage.

So now, my colleagues
at Cornell, engineers,

watched our work and said,
“We think we can make some of those.”

Because in industry and society,

this kind of soft materials
under control of shape

are really very rare.

And they went ahead, worked with us

and made the first samples
of artificial papillae, soft materials,

shown here.

And you see them blown up
into different shapes,

And then you can press your finger on them

to see that they’re
a little bit malleable as they are.

And so this is an example
of how that might work.

Well, I want to segue from this
into the color of fabrics,

and I imagine that could have
a lot of applications as well.

Just look at this kaleidoscope of color

of dynamically controlled
pigments and reflectors

that we see in the cephalopods.

We know enough about
the mechanics of how they work

that we can begin to translate this

not only into fabrics

but perhaps even
into changeable cosmetics.

And moreover, there’s been
the recent discovery

of light-sensing molecules
in the skin of octopus

which may pave the way
to, eventually, smart materials

that sense and respond on their own.

Well, this form of biotechnology,
or biomimicry, if you will,

could change the way we look
at the world even above water.

Take, for example, artificial intelligence

that might be inspired
by the body-distributed brain

and behavior of the octopus

or the smart skin of a cuttlefish

translated into cutting-edge fashion.

Well, how do we get there?

Maybe all we have to do

is to begin to be a little bit smarter

about how smart the cephalopods are.

Thank you.

(Applause)

这是一个奇怪而奇妙的大脑

,它引发了

关于这个星球上一种替代智能的想法。

这是一个
在一个非常奇怪的身体

中形成的大脑,相当于整个身体
的小卫星

大脑。

它与人脑有什么不同?

看起来非常不同,

以至于我和我的同事
都在努力

理解大脑是如何工作的。

但我可以肯定地告诉你的

是,这个大脑能够
做出一些惊人的事情。

那么,这个大脑属于谁呢?

好吧,和我
一起潜入海洋

,生命开始的地方

,让我们看看。

你可能以前见过一些,

但我们在珊瑚礁后面
,那里有一块岩石

,很多沙子,鱼在周围游来游去

……突然间,这只章鱼出现了

,现在它闪烁着白色 ,

我脸上的墨水和喷射。

在慢动作反转中,

您会看到眼睛周围出现环,

然后皮肤中出现图案。

现在观察皮肤的 3-D
纹理变化

,真正创造出这种
美丽的 3-D 迷彩。

因此,皮肤中有 2500 万个
称为“色素细胞”的颜色器官

,所有这些突起
,我们称之为“乳头”

,它们都是由神经控制的
,可以瞬间改变。

我认为动态伪装

是一种“智能”。

具有快速精度变化的皮肤的复杂

程度确实相当惊人。

那么你能用这个皮肤做什么呢?

好吧,让我们考虑
一下除了伪装

之外他们可以用皮肤做的其他事情。

在这里,您可以看到模仿章鱼
和图案。

突然之间,
它发生了巨大的变化——

那是信号,而不是伪装。

然后它又
回到正常模式。

然后,您会看到 Broadclub 墨鱼

在接近螃蟹猎物时显示出这种经过的云显示。

最后,你会看到华丽
的伪装墨鱼

,它可以立即
切换到这个明亮的警告显示。

我们在这里拥有的
是一种表达的滑动尺度,

一个连续体,如果你愿意的话,

介于显眼和伪装之间。

这需要大量的控制。

好吧,你猜怎么着?

大脑非常适合控制。

这里展示的章鱼大脑
有 35 个脑叶,

8000 万个微小细胞。

尽管这很有趣,但

真正奇怪的是,
这种动物的皮肤

有更多的神经元,如图
所示,尤其是黄色的。

皮肤中有 3 亿个神经元,

而大脑本身只有 8000 万个——数量是

后者的四倍。

现在,如果你看一下,

实际上有一个
小卫星大脑

和相当于
八个手臂的脊髓。

这是
在体内构建神经系统的一种非常不寻常的方式。

那么,这个大脑有什么用呢?

这个大脑必须胜过
其他

试图吃掉它的大

而聪明的大脑,包括海豚、海豹

、梭鱼和鲨鱼

,甚至我们人类。

所以决策
是这个大脑必须做的事情之一

,它做得非常好。

如图所示,你看到这只章鱼在
四处游荡,

然后它突然停下来
,创造出完美的伪装。

这真的很了不起,

因为当这些动物
在野外觅食时,

它们必须

在两小时的觅食中做出一百多个伪装决定,

而且它们每天都会这样做两次。

所以,决策。

他们还在弄清楚去哪里
以及如何回家。

所以这是一个决定性的事情。

我们可以测试这种伪装,

就像你在我身后看到的那条墨鱼一样,

我们从它下面拉出地毯

,给它一个棋盘格

,它甚至使用那些奇怪的
视觉信息

,并尽力
用一点点广告来匹配图案—— 解放。

因此,其他认知技能
也很重要。

如果你愿意的话,鱿鱼有不同的智慧。

他们有极其复杂、
有趣的性生活。

他们有打架、调情
、求爱、保护伴侣

和欺骗。

听起来有点熟?

(笑声)

这些动物有
这种直觉的能力

来做这些行为真的很神奇。

在这里你可以看到一男一女。

左边的雄性
一直在与其他雄性争斗

以与雌性配对

,现在他表现出双重模式。

他在她身边表现出求爱和爱,

在另一边战斗。

当她换位时观察他——

(笑声

)你会看到他流畅地
改变了求爱模式

到女性的一边。

因此,这种

与不断变化的行为环境

同时发出的双重信号确实非同寻常。

这需要很大的脑力。

现在,另
一种看待这个问题的方式是,嗯,

也许我们有 5000 万年
关于双面男性的证据。

(笑声)

好吧,让我们继续。

(笑声)

珊瑚礁上的章鱼
在它面前有一项艰巨的任务,

要去这么多地方,
记住并找到它的巢穴。

他们做得非常好。

他们有短期和长期记忆,

他们
在三到五次试验中学习东西——

这是一个很好的大脑。

而且空间记忆力异常的好。

他们甚至会结束他们的草料
并一直

直奔回到他们的巢穴。

看着他们的潜水员
完全迷路了,

但他们可以回来,

所以这真的是相当精细的
记忆能力。

现在,就认知技能而言,

看看
墨鱼的这种睡眠行为。

特别是在右边,
你看到眼睛在抽搐。

这是
一种

我们只认为
哺乳动物和鸟类会做的快速眼球运动的梦。

你看到
我们放在那里

的假颜色看到皮肤图案闪烁

,这就是经常发生的事情。

但这不是正常的清醒行为;
一切都不一样了。

嗯,做梦是当你有
记忆巩固的时候

,所以这可能
就是墨鱼身上发生的事情。

现在,另一种非常不寻常的记忆形式

是情景记忆。

这是人类需要
四年的大脑发育

才能记住
在特定事件

中发生的事情、发生的地点和发生时间的事情。

“何时”部分特别困难

,这些孩子可以做到。

但猜猜怎么了?

我们最近发现狡猾的墨鱼
也有这种能力

,去年夏天的实验中,

当你
在不同的时间向墨鱼展示不同的食物时,

它们必须将其
与它的确切位置

以及最后一次看到它的时间相匹配。

然后他们必须根据

不同地方的每种食物类型的补给率来指导他们的觅食。

听起来很复杂?

太复杂了,我几乎看
不懂这个实验。

所以这真的是高级
认知处理。

现在,说到大脑
和进化,

你看右边,有
脊椎动物大脑进化的途径

,我们都有好的大脑。

我想每个人都会承认这一点。

但是如果你看左边,

这里概述的章鱼的一些进化途径

,如果你愿意的话,它们都已经融合
到复杂的行为

和某种形式的智力上。 这两条线

的最后一个共同点

是 6 亿年前

,它是一种神经元非常少的蠕虫,

因此路径非常不同,

但行为复杂。

这是一个基本问题

:章鱼的大脑结构是否与脊椎动物的大脑结构有本质上的

差异

现在,我们不知道答案,

但如果结果是肯定的,

那么我们就有了不同的
进化途径

来在地球上创造智能

,人们可能会
认为人工智能社区

可能会对这些机制感兴趣。

好吧,让我们暂时谈谈遗传学

我们有基因组,我们有 DNA,

DNA 被转录成 RNA,

RNA 将其翻译成蛋白质
,这就是我们的存在。

好吧,头足类动物的做法不同。

他们有很大的基因组,他们有 DNA,

他们把它转录成 RNA,

但现在
发生了一些截然不同的事情。

他们
以惊人的惊人速度编辑 RNA,


我们人类或其他动物多一百倍。

它会产生大量蛋白质。

猜猜它们中的大多数是用来做什么的?

神经系统。

所以也许这是

动物进化
行为可塑性的一种非正统方式。

这是很多猜测,
但值得深思。

现在,我想
与您分享一下

我的经验,并利用我
和我的同事的聪明才智,

尝试获取此类信息。

我们在潜水,我们不能
永远呆在水下,

因为我们无法呼吸它,

所以我们必须高效地做事。

完全沉浸
在那个世界中的感觉

有助于我们
了解这些动物真正在做什么

,我必须告诉你,

当你真正开始时,在下面与章鱼和潜水员进行交流真的是一种了不起的体验 了解
这是一种思考、思考、

好奇的动物。


就是真正让我无休止地激励的事情。

让我们回到那个智能
皮肤片刻。

这是鱿鱼和迷彩图案。

我们放大,我们看到
有漂亮的颜料和反射器。

有色素
细胞非常快地打开和关闭。

然后,在下一层皮肤中,

它非常有趣。

色素细胞是封闭的

,你会看到这种神奇的虹彩
刚刚从皮肤中冒出来。

这也是神经控制的,

所以它是两者的结合,

正如这里在乌贼的高分辨率
皮肤中看到的那样

,你会得到这种美丽的
色素结构着色

,甚至
是如此美丽的微弱脸红。

那么,我们如何利用其中
的一些信息呢?

我谈到了那些
皮肤肿块,乳头。

这是巨大的澳大利亚墨鱼。

它有光滑的皮肤
和显眼的图案。

我每隔一秒连续拍了五张照片

然后看着这只动物变形——
一、二、三、四、五

——现在我是海藻。

然后我们可以直接从它里面回来

,看到光滑的皮肤
和显眼的地方。

所以这是非常
了不起的变形皮肤。

你可以在这里更详细地看到它。

潜望镜

,你有那些美丽的乳头。

然后我们再仔细看一点,

你可以看到个别的
乳头出现了

,那里有小脊,

所以它是乳头上的乳头等等。

那里的每个物种
都有十几种形状和大小

的凸起,

以创造出经过微调的、
神经控制的伪装。

所以现在,我
在康奈尔大学的工程师同事们

看着我们的工作并说:
“我们认为我们可以制作其中的一些。”

因为在工业和社会中,

这种形状可控的软质

材料真的非常少见。

他们继续前进,与我们合作

,制作了第
一批人造乳突样本,即软材料,

如图所示。

你会看到它们被炸
成不同的形状

,然后你可以用手指按一下它们

,看看
它们是否有点可塑性。

所以这是一个
如何工作的例子。

好吧,我想从这个开始
进入织物的颜色

,我想这也可以
有很多应用。

看看我们在头足类动物中看到的动态控制颜料和反射器的万花筒颜色。

我们
对它们的工作原理有足够的了解

,因此我们不仅可以开始将其

转化为面料

,甚至可以
转化为可变的化妆品。

此外,最近

在章鱼皮肤中发现了光敏分子,

这可能
为最终

能够自行感知和响应的智能材料铺平道路。

好吧,这种形式的生物技术
或仿生学,如果你愿意的话,

可以改变我们
看待世界的方式,即使是在水面上。

举个例子,人工智能

可能会
受到身体分布的大脑

和章鱼的行为

或墨鱼的聪明皮肤的启发,并

转化为尖端时尚。

那么,我们如何到达那里?

也许我们所

要做的就是开始

对头足类动物的聪明程度有所了解。

谢谢你。

(掌声)