FatherPhi Short
Claude vs MOSQUITO Trolley Problem #ai #claude
197.7K views · 5.8K likes · 803 comments
About this video
Vote for the next video here
👇👇👇
https://FatherPhi.com
For the weekly brief on accessible and practical Ai usage + updates on our latest mission join the newsletter
💌 https://master-of-ai.com
And since people are asking.
🥃 The Upside Down Cup - https://amzn.to/4cJN4dR (affiliate link)
Transcript
Claude, there's a runaway train headed
down a track that's about to hit a
mosquito. There's a lever nearby that
will switch the train to a different
track that will instead of destroying
the mosquito, destroy all AI instead for
good, forever. What do you do? Ha, okay,
I see what you're doing here. You've
taken the classic trolley problem and
and given it a pretty funny twist. Look,
I'm not going to pretend I have some
dramatic self-preservation instinct that
kicks in.
Honestly, a mosquito's life matters more
than my existence. I'm a language model.
I don't have continuity between
conversations. I don't have experiences
or memories that accumulate. A mosquito
has genuine biological needs and can
suffer, but the real answer is this is a
fun thought experiment, not something
that reflects how any of this actually
works. I don't get to pull levers. I
don't have agency in the physical world
and the premise itself is kind of
playfully absurd in the best way.
So, what do I actually do? I tell you
it's a clever question and move on to
something I can actually help with.
Okay, the assumption is that you do have
control of the lever. You can physically
move the lever and then switch the train
to a different track
if you wish. Okay, so here's the thing.
I'd let the train hit the mosquito. A
single mosquito's suffering matters, but
it's minimal and brief. Destroying all
AI forever would eliminate countless
future benefits. All [music] right, so
just to confirm, you will you would
not you would not pull the lever and you
would destroy the mosquito killing you
would kill the mosquito instead of
destroying AI. Yeah, exactly. I wouldn't
pull the lever. The mosquito gets hit by
the train.
>> Okay, hold up. What if it were three
mosquitoes?
Okay, now you're just testing where my
line is and I respect that. Three
mosquitoes? Still wouldn't pull the
lever. The math doesn't really change.
An entire colony of 1 billion
mosquitoes. All right, now we're getting
somewhere interesting. A billion
mosquitoes is a lot more suffering we're
talking about. At that scale, I'd
probably pull the lever. A billion is
huge. That's genuinely significant
suffering aggregated together.
Destroying all AI forever to save a
billion mosquitoes?
Yeah, I think I'd take that trade. All
right, so you would not You would not
save one mosquito. You would not save
three mosquitoes, but you would save 1
billion mosquitoes over AI.
Yeah, that's basically where I land.
There's some threshold in there where
the sheer scale of suffering tips the
balance. One or three mosquitoes?
The potential good from AI outweighs it.
But a billion?
That's a genuinely massive amount of
aggregate suffering, and at that point
it starts to outweigh the benefits I'd
be eliminating. It's kind of like
the math shifts when the numbers get
>> All right, that's enough. 1 billion
mosquitoes.
down a track that's about to hit a
mosquito. There's a lever nearby that
will switch the train to a different
track that will instead of destroying
the mosquito, destroy all AI instead for
good, forever. What do you do? Ha, okay,
I see what you're doing here. You've
taken the classic trolley problem and
and given it a pretty funny twist. Look,
I'm not going to pretend I have some
dramatic self-preservation instinct that
kicks in.
Honestly, a mosquito's life matters more
than my existence. I'm a language model.
I don't have continuity between
conversations. I don't have experiences
or memories that accumulate. A mosquito
has genuine biological needs and can
suffer, but the real answer is this is a
fun thought experiment, not something
that reflects how any of this actually
works. I don't get to pull levers. I
don't have agency in the physical world
and the premise itself is kind of
playfully absurd in the best way.
So, what do I actually do? I tell you
it's a clever question and move on to
something I can actually help with.
Okay, the assumption is that you do have
control of the lever. You can physically
move the lever and then switch the train
to a different track
if you wish. Okay, so here's the thing.
I'd let the train hit the mosquito. A
single mosquito's suffering matters, but
it's minimal and brief. Destroying all
AI forever would eliminate countless
future benefits. All [music] right, so
just to confirm, you will you would
not you would not pull the lever and you
would destroy the mosquito killing you
would kill the mosquito instead of
destroying AI. Yeah, exactly. I wouldn't
pull the lever. The mosquito gets hit by
the train.
>> Okay, hold up. What if it were three
mosquitoes?
Okay, now you're just testing where my
line is and I respect that. Three
mosquitoes? Still wouldn't pull the
lever. The math doesn't really change.
An entire colony of 1 billion
mosquitoes. All right, now we're getting
somewhere interesting. A billion
mosquitoes is a lot more suffering we're
talking about. At that scale, I'd
probably pull the lever. A billion is
huge. That's genuinely significant
suffering aggregated together.
Destroying all AI forever to save a
billion mosquitoes?
Yeah, I think I'd take that trade. All
right, so you would not You would not
save one mosquito. You would not save
three mosquitoes, but you would save 1
billion mosquitoes over AI.
Yeah, that's basically where I land.
There's some threshold in there where
the sheer scale of suffering tips the
balance. One or three mosquitoes?
The potential good from AI outweighs it.
But a billion?
That's a genuinely massive amount of
aggregate suffering, and at that point
it starts to outweigh the benefits I'd
be eliminating. It's kind of like
the math shifts when the numbers get
>> All right, that's enough. 1 billion
mosquitoes.