Stop Asking, Start Doing
(shortpost on AI)
Duncan Sabien
Sep 04, 2026
52
31
5
Share
“Okay, but like. What am
I
supposed to do about it?”
I hear this all the time, from people who have just spent the last 5 or 10 or 30 minutes learning about the state of AI development, and how incredibly dangerous and unstable the situation is.
(If you’re out of the loop: we could actually for-real no-joke all die within the next 5 years. I am not exaggerating and I am not a kook. It’s that bad. And most of the worlds where we don’t die in the next 5 years are worlds where we die in the next 15. You will not live to die of climate change if we don’t do something about AI
now.
)
And the response—“what do you want
me
to do?”—always feels
off,
to me. There is something deeply wrong with it.
Imagine for a moment that there is a 100km meteor headed for Earth. A literal actual giant hunk of rock, not a
metaphorical
meteor. The kind that will wipe out 100% of all life, that would not be survivable even if you have a bunker and tons of food because it will melt the surface and ignite the atmosphere.
Additionally,
imagine that the meteor has some strange property, such that if 51+% of people all agree to push a button, it will simply disappear. It’s the kind of life-ending threat that we can
literally
just vote away.
I think that most people, in that situation, would
not
go “Okay, but like. What do you want
me
to do about it?”
I want you to agree to push the button! And to help get everybody else to agree to push the button! And there are like a
million
different ways to help with that.
If you have lots of money, then pay people to push the button, or fund pro-button journalism. If you have lots of friends, tell
all of them
about the button, and ask them all to push it. If you know bright clever young people who are good at coming up with wild and crazy ideas that sometimes pan out big, then set them to work on the hard problem of making button-pressing popular.
“Okay, but what do you want
me
to do about this?” are the weasel words of a person who does not, in fact, understand that they will
die
when this meteor hits. There are indeed
some
people who, when faced with a hard and lethal challenge with slim odds of success, will just roll over and give up. There are
some
people who are tired, and defeatist, and are just like “welp, guess I’ll die, then.”
But
most
people are not like that. Most people will scrabble and scramble and fight tooth and nail for the slim chance of survival…
…if they in fact understand that they are facing death. If they have comprehended the situation fully, and are taking the problem seriously.
When someone says to me “okay sure I buy that AI is scary and dangerous and might kill literally everyone and everything,
soon
even, but like, what do you want
me
to do about it?” what I think is usually going on is that they don’t actually believe it. They haven’t actually grokked it. They’re flinching or looking away, on some level.
It’s one thing if a person is genuinely asking. Like, some people have the vibe of “okay, wow, I did not know about all this, this was news to me, thank you for telling me. And
now that you’ve told me
, do you happen to have an existing, structured effort that I can just … plug in to? I mean, you seem to know what’s going on, here—can I just pitch in? Are you ready for me?”
That
seems totally reasonable, as a response.
But when the answer comes back “no, sorry, this whole thing is a scrappy scramble, there are a few dozen organizations working on various tacks but in fact nobody has a solid plan for getting 51% of people to push the button,” then the person who really truly gets it, and really truly understands the problem, does not say “oh, well, I guess that means there’s nothing for
me
to do, here. Guess I just go back to my day job and try not to think about it. Too bad you don’t have a ready-made and clearly defined task for me.”
No! Fucking no!
To be clear: you may be
limited
in your ability to help. You may not have
lots
that you can do. You may need to keep your day job, because there isn’t actually remunerative sustainable work for you in trying to tackle this problem full-time.
But
you can help a little even if you can’t help a lot.
You can do things here and there while continuing to live a life that is as good as you can make it, in the shadow of the oncoming meteor.
“But all of my friends already know, so it’s not like I can help with the whole awareness problem—”
Bullshit.
You have parents, and grandparents, and aunts and uncles. You have neighbors. You have cousins and nieces and nephews and coworkers. You have friends, who have friends, who have friends. Some of them are politicians, or journalists, or know people in the tech industry, or know people who know people who know billionaires.
And they’ll all die if the meteor hits.
They will
all die
if we build superintelligent AI under the current paradigm, which the people at OpenAI and Anthropic
are trying to do
, and they’re making progress
fast
.
We need 51% of people to push the button.
If we get 51% of people to push the button, the meteor disappears.
This is amazing. This makes it the
easiest
extinction challenge we’ve ever faced. Literally all we have to do is not build the giant meteor-attracting magnet!
And you can help with that. Stop asking “but what am
I
supposed to do?” and act as though you take the problem
seriously
. And if you
don’t
take the problem seriously—if you are not convinced—then say
that.
It’s fine to be skeptical! There
are
lots of historical examples of people claiming the world would end, and being wrong about that. We’re well aware of the narrative awkwardness of the situation, and used to dealing with doubt and suspicion. We will
happily
drown you in evidence showing the overwhelming case for concern.
But don’t accept the evidence, agree with the conclusion, and then throw up your hands as if there’s nothing you can do.
That
is
crazy.
Everything I write becomes free after a week or two, but it’s only paid subscriptions that make it
possible
for me to write. If you found a coffee’s worth of value in this or any of my other writing, please consider signing up to support me; every bill I can pay with writing is a bill I don’t have to pay by doing other stuff instead. I also accept and greatly appreciate one-time donations of any size at
paypal.me/duncansabien
52
31
5
Share
What did you think?
Your reaction is separate from the prediction. It won’t change your rule.