0:00

Happy Monday! I'm Michael Kavnatt, host

0:02

of The Next Big Idea Spinoff,

0:04

The Next Big Idea Daily. Every

0:06

weekday, I invite leading nonfiction authors

0:08

to share the big ideas from

0:10

their books about psychology, business, relationships,

0:12

and just about everything else. It's

0:14

a mini master class you can

0:16

listen to in the time it

0:18

takes to walk your dog. You're

0:20

about to hear the episode we

0:22

published this morning. To get the

0:24

rest of this week's episodes, follow

0:26

the next Big Idea Daily, wherever

0:28

you get your podcasts. LinkedIn

0:30

Presents. Good morning. I'm

0:32

your host Michael Kavnath and

0:35

this is the next big idea

0:37

daily. Now lately there's been

0:40

a lot of debate about

0:42

artificial intelligence. Is it going

0:44

to destroy humanity or save

0:47

it? Is it really capable

0:49

of thinking? of consciousness. But

0:51

perhaps the most provocative question

0:53

we can ask is, does

0:56

AI have rights? What responsibilities

0:58

do we have to these

1:00

increasingly sophisticated systems? These questions

1:03

might sound futuristic, but Jeff

1:05

Sibo, a leading thinker in

1:07

environmental ethics, argues we've been avoiding

1:09

them for too long, not just

1:12

for AI, but for animals, plants,

1:14

and even entire ecosystems. In his

1:16

new book, The Moral Circle, Who

1:18

matters, what matters, and why? Jeff

1:20

makes the case that we need

1:22

to expand our ethical concerns beyond

1:24

humans. If there's any chance that

1:26

AI or a fish or an

1:28

octopus or even a fungus has

1:30

sentience, agency or moral significance, then

1:32

we should act as if they

1:34

do. Because the way we treat

1:37

them today will shape the kind of

1:39

world we live in tomorrow. Here's Jeff

1:41

to share five big ideas from his

1:43

book right after this quick break. If

1:52

you're interested in the story behind

1:55

the business headlines, check out big

1:57

technology podcast. My weekly show that

1:59

features in-depth interviews. with CEOs, researchers,

2:01

and reformers in business and technology.

2:04

Hi, I'm Alex Kantruitz. I'm a

2:06

long-time journalist, CMBC contributor, and the

2:08

host of the show. I emptied

2:10

my Rolex every Wednesday to bring

2:13

you awesome episodes, so go check

2:15

out Big Technology Podcast. It's available

2:17

on all podcast apps. I'd love

2:19

to have you as listener. Hi,

2:26

my name is Jeff Sivo and

2:28

I direct the Center for Environmental

2:30

and Animal Protection and the Center

2:33

for Mind Ethics and Policy at

2:35

New York University. And I'm here

2:37

today to tell you about my

2:39

book, The Moral Circle, Who Matters,

2:42

What Matters, and Why. We share

2:44

the world with a vast number

2:46

and wide range of non-human. That

2:48

includes other animals, vertebrates and invertebrates.

2:50

It includes other living beings, plants,

2:53

plants, plants, plants. And in all

2:55

these cases, we have disagreement and

2:57

uncertainty about which beings matter and

2:59

what we owe them. So how

3:01

can we make decisions together about

3:04

how to interact with such a

3:06

vast number and wide range of

3:08

non-human despite substantial ongoing disagreement and

3:10

uncertainty about these ethical and scientific

3:12

issues? That is the question this

3:15

book addresses. So we can hear

3:17

focus on five key ideas that

3:19

come out in the book, starting

3:21

with idea number one, if you

3:23

might matter. we should assume that

3:26

you do. In ethics, we have

3:28

a lot of disagreement and uncertainty

3:30

about what it takes to matter.

3:32

Some people think you need to

3:34

be sentient, able to consciously experience

3:37

pleasure and pain. Other people think

3:39

you need to be agentic, able

3:41

to set and pursue your own

3:43

goals based on your own beliefs

3:45

and desires. Other people think you

3:48

need to be alive, able to

3:50

perform basic life functions associated with

3:52

survival and reproduction. Now in science

3:54

we also have disagreement and uncertainty

3:56

about which beings have these features

3:59

with other mammals. and birds, we

4:01

can be confident based on the

4:03

information currently available, that they are

4:05

not only alive, but also sentient

4:07

and agentic and morally significant. They

4:10

can feel, they can think, and

4:12

they matter for their own sakes.

4:14

But what about the other vertebrates,

4:16

reptiles, amphibians, fishes? What about... invertebrates

4:18

with more distributed cognitive systems? What

4:21

about plants and fungi with radically

4:23

different kinds of cognitive systems? What

4:25

about chatpots and robots that are

4:27

made out of silicon instead of

4:29

being made out of carbon? In

4:32

these cases, we might genuinely be

4:34

uncertain about whether it feels like

4:36

anything to be them, for example,

4:38

for a long time. Well, I

4:40

argue that we should not wait

4:43

for certainty before taking basic steps

4:45

to treat these non-human as well.

4:47

If there is at least a

4:49

realistic non-negligible chance that they matter

4:51

for their own sakes, based on

4:54

the best information and arguments currently

4:56

available, then we should take reasonable,

4:58

proportionate steps to consider and mitigate

5:00

the risks that our actions and

5:02

policies might be imposing on them.

5:05

Many beings might matter. How can

5:07

we tell which non-human or sentient

5:09

or agentic or otherwise morally significant?

5:11

Well, when proof and certainty are

5:13

unavailable, we can at least collect

5:16

evidence and estimate probabilities. And in

5:18

particular, we can use a marker

5:20

or indicator method to assess non-humans

5:22

for behavioral or anatomical evidence associated

5:24

with capacities like sentience and agency.

5:27

With animals, we can use behavioral

5:29

tests. We can ask, for example,

5:31

do they nurse their own wounds?

5:33

Do they respond to analgesics and

5:35

antidepressants in the same ways as

5:38

us? Do they make behavioral tradeoffs

5:40

between the avoidance of pain and

5:42

the pursuit of other valuable goals?

5:44

To the extent that the answer

5:46

is yes, it increases the probability

5:49

of moral significance. And when we

5:51

ask these questions about animals, the

5:53

answer is often yes. In fact,

5:55

at this point, many experts in

5:58

many fields are prepared to say

6:00

that there is at least a

6:02

realistic not- negligible chance of moral

6:04

significance in all vertebrates, mammals, birds,

6:06

reptiles, amphibians, and fishes, and many

6:09

invertebrates. Cephalopod mollus like octopuses, decapod

6:11

crustaceans, like lobsters, even insects, like

6:13

ants, and bees. Now with AI

6:15

systems, we might not be able to trust

6:17

behavioral evidence in the same kinds of

6:20

ways of present, but we can look

6:22

past potentially misleading misleading behaviors

6:24

at... underlying architectures, and we

6:27

can ask, do AI systems

6:29

have computational features that we

6:31

associate with capacities like sentience and

6:33

agency? For example, do they have

6:36

their own forms of perception, attention,

6:38

learning, memory, self-awareness,

6:40

social awareness, language, language, and

6:43

reason? To the extent that

6:45

the answer is yes, that once

6:47

again increases the probability of moral

6:49

significance. And while current AI

6:51

systems might not have many of these

6:53

capacities at all, We can expect

6:56

that near future AI systems will

6:58

have advanced and integrated versions of

7:01

many of these capacities, but they

7:03

happen to be built out of silicon

7:05

instead of carbon. And so I argue

7:07

that we should give at least minimal

7:09

moral consideration to all vertebrates,

7:12

many invertebrates, and many near

7:14

future AI systems in the

7:16

spirit of caution and humility

7:19

based on the best information

7:21

and arguments currently available. If

7:24

we might be affecting you, we should assume

7:27

that we are. In addition to having disagreement

7:29

and uncertainty about which beings

7:31

matter, we also have disagreement

7:33

and uncertainty about what we

7:35

owe everyone who matters. In ethics,

7:37

we still debate whether we have

7:39

a general responsibility to help each

7:42

other. Some people think that we

7:44

should. If I can prevent something

7:46

very bad from happening without sacrificing

7:48

anything comparably significant, then I should

7:51

help. Other people think, no, we do not

7:53

have a general responsibility to help

7:56

others. Yes, I should consider the risks that

7:58

I might be imposing on a... and I

8:00

should reduce and repair the harms

8:02

that I cause to others, but

8:04

beyond that, helping is optional, not

8:06

required. And then in science, we

8:09

often have disagreement and uncertainty about

8:11

whether our actions and policies are

8:13

in fact imposing risks and harms

8:15

on vulnerable others. Suppose you dump

8:18

a bunch of toxic untreated waste

8:20

in a lake, and then suppose

8:22

you walk by that lake the

8:24

next day, and you see a

8:27

rabbit drowning in the middle of

8:29

the lake. Did you play a

8:31

role in this predicament? Well, you

8:33

might not have directly imperiled the

8:36

rabbit. You might not have picked

8:38

her up and plopped her in

8:40

the middle of the lake. But

8:42

you might have indirectly imperiled her.

8:45

Your toxic untreated waste might have

8:47

played a role in her getting

8:49

stuck in the middle of the

8:51

lake. So I argue that we

8:54

should cultivate caution and humility in

8:56

the face of disagreement and uncertainty

8:58

about these ethical and scientific issues

9:00

as well. In this case, for

9:03

example, you should help the rabbit,

9:05

either because you might have a

9:07

responsibility to help others where possible,

9:09

or at least because your own

9:12

actions might have indirectly imperiled this

9:14

rabbit, and so helping her as

9:16

a way of reducing and repairing

9:18

the harms that you are personally

9:21

causing in the world. But if

9:23

we do have these responsibilities, then

9:25

we have to ask, how often

9:27

do they arise? How often are

9:30

we in a position of, if

9:32

not helping, then at least reducing

9:34

and repairing the harms that we

9:36

cause to vulnerable others? We might

9:39

be affecting many beings. We now

9:41

live in the Anthropocene, a geological

9:43

epoch where humanity is a dominant

9:45

influence on the planet. We are

9:48

now affecting non-humans all over the

9:50

world, whether we like it or

9:52

not, both directly. and indirectly, both

9:54

individually and collectively. Consider industrial animal

9:57

agriculture. This food system kills hundreds

9:59

of billions of captive vertebrates and

10:01

trillions of captive invertebrates for food

10:03

every year. To say nothing of

10:06

all the wild animals killed for

10:08

food every year, to say nothing

10:10

of all the animals killed for

10:12

other purposes every year. And this

10:15

food system also very significantly increases

10:17

global health and environmental threats, including

10:19

threats associated with the spread of

10:21

diseases, antimicrobial resistance, pollution, biodiversity loss,

10:24

human-caused climate change. And then when

10:26

these threats occur, they imperil humans

10:28

and non-humans alike, both directly and

10:30

indirectly. They imperilice directly by exposing

10:33

us to diseases and fires and

10:35

floods, and they imperilice indirectly by

10:37

amplifying. ordinary threats that we already

10:39

face, like threats associated with hunger,

10:42

thirst, illness, injury. For animals, they

10:44

amplify threats associated with human violence

10:46

and neglect. And in the future,

10:48

we can expect similar dynamics to

10:51

occur with emerging technologies like the

10:53

development and deployment of advanced AI

10:55

systems. If and when AI systems

10:57

are sentient, agentic or otherwise morally

11:00

significant, we could be using them

11:02

at even greater scales than we

11:04

do with nonhuman animals right now.

11:06

And then when we do that,

11:09

we could also be creating and

11:11

amplifying a wide range of threats,

11:13

we could lose control of AI,

11:16

and AI could harm us. We

11:18

could retain control of AI, and

11:20

we could use AI to harm

11:22

each other. And, AI might amplify

11:25

ordinary threats that we already face,

11:27

like threats associated with bias, disinformation,

11:29

and even graver threats in the

11:31

future. In all of these cases,

11:34

I argue that we have a

11:36

responsibility to consider all affected stakeholders

11:38

equitably when making decisions about our

11:40

effects in the world. That includes

11:43

humans, animals, and eventually, AI systems.

11:45

exceptionalism, the presumption that humanity always

11:47

matters most and takes priority. If

11:49

such a vast number and wide

11:52

range of non-humans might matter, and

11:54

if our actions and policies might

11:56

be affecting them at global scales,

11:58

then we owe them a lot.

12:01

Now many humans assume that we

12:03

should nevertheless prioritize fellow humans because

12:05

we have higher capacities for welfare.

12:07

I can suffer more than a

12:10

mouse, for example. But first of

12:12

all, we might not... always have

12:14

higher capacities for welfare, I might

12:16

not be able to suffer more

12:19

than an elephant or a whale

12:21

or in the future a very

12:23

sophisticated AI system. And second of

12:25

all, even if we have higher

12:28

capacities for welfare than non-human's individually,

12:30

we might not have higher capacities

12:32

for welfare than them in the

12:34

aggregate because the non-human population is

12:37

and will be much larger and

12:39

much more diverse than the human

12:41

population. They have more at stake

12:43

than we do overall, even if

12:46

we have more at stake than

12:48

they do individually. Now some humans

12:50

also assume that we should prioritize

12:52

fellow humans because we have closer

12:55

bonds with fellow humans. But that

12:57

might not always be true either.

12:59

If we really are affecting non-human

13:01

everywhere whether we like it or

13:04

not, then we have morally significant

13:06

bonds with them too. We are...

13:08

impacting them, and we have a

13:10

responsibility to reduce and repair the

13:13

harms that we are causing them.

13:15

So I think that we might

13:17

not be able to sustain our

13:19

assumption that we always take priority.

13:22

Now there might be a limit

13:24

to how much we can support

13:26

non-human right now because we lack

13:28

the knowledge and power and political

13:31

will that we need to help

13:33

them. But we can still do

13:35

more than we are at present,

13:37

and we can also make an

13:40

effort to build knowledge, build capacity,

13:42

build political will, towards helping them

13:44

in the future. And if and

13:46

when we have the ability to

13:49

prioritize them in an effective and

13:51

sustainable way, perhaps we should prioritize

13:53

them at that point. Now I

13:55

know that some of these ideas

13:58

might seem implausible. They might even

14:00

seem like a distraction from the

14:02

other very important issues that we face

14:04

right now. But I think that we

14:07

should take them seriously anyway alongside those

14:09

other very important issues. As the dominant

14:11

species, we have a responsibility

14:13

to ask how we might

14:15

be affecting all stakeholders, including

14:17

humans and animals and eventually

14:19

potentially AI systems. And if that

14:21

leads us to Uncomfortable conclusions about our

14:24

treatment of non-human's, so be it. We

14:26

can accept those conclusions and we can

14:28

try to build a better world for

14:30

humans and non-human's alike. Thank

14:34

you Jeff. Okay, listeners, you can get a

14:37

copy of The Moral Circle wherever you get

14:39

your books. And make sure you're subscribed to

14:41

the next Big Idea Daily Feed because this

14:44

week we'll be talking about what climate change

14:46

is doing to your health, how soon we

14:48

could be mining on Mars, and a whole

14:50

lot more. I'm Michael Kavnat. I hope to

14:53

see you tomorrow.