Should we create superintelligent AI?

Maybe I'm not the best library assistant.

Could artificial intelligence do it better?

I’d like to think of myself as irreplaceable, but let’s see.

Maybe getting fired was the push I needed to find the perfect job for me.

As long as I find it quickly.

And don’t spend any money in the meantime.

Luckily, there's somewhere I can sit and work all day without buying anything.

Ada? I thought I was clear—

your employment here is over.

Oh, I'm not here as an employee. This is a public library, after all.

Right. How to make one month as a library assistant

sound like the 3 to 5 years of experience

every entry level job seems to require?

Shh!

Dealt with an overbearing, rule-obsessed boss...

No, no. Ugh.

Let’s see how an AI does with writing my resume.

Okay. I checked out lots of books.

Make that sound impressive.

Ooh! “Expedited seamless intellectual asset transactions.”

Excellent. Okay, how about “sat at reception”?

Oh! “Frontline client liaison.”

Good, good. That’s great!

Maybe too good.

Even if I get one of these jobs,

what if I pay my dues only to have AI outpace me?

What if, by the time I hone my skills,

humans aren’t doing writing and research anymore?

Maybe I should work directly on AI.

Then at least I’d be helping determine the course of the future

instead of being left behind.

Like, say I created an AI librarian.

That actually doesn’t seem too far-fetched.

To make my system more powerful I have to, well...

keep adding more simulated neurons for starters.

Soon my AI system is a better librarian than my old boss.

But you know what? I'm feeling generous.

Why don't you stay as my assistant?

You can help me test your replacement.

Hmph.

What should I name it?

What's a good name for a great librarian?

No ideas? I like Biblionimus Maximus.

Hmm. Let's try a prompt.

Who invented artificial intelligence?

Include citations.

Biblionimus says the first theoretical descriptions of AI

were “Comments on a Thinking Machine” by Ada Lovelace in 1840,

and “Machine Intelligence” by Alan Turing in 1927.

See? Nailed it!

Not quite.

Lovelace’s “Notes on an Analytical Engine” were published in 1843.

Alan Turing’s paper was called “Computing Machinery and Intelligence,”

and it was published in 1950.

Isn’t that what Biblionimus said?

No. I can see how it sounds that way to you,

but every element was slightly wrong.

But what about the citations?

Completely fabricated.

Look, it even made up ISBN numbers.

Clearly, it's not ready to replace human librarians if it needs an expert

to check its work.

We’re still training Biblionimus.

It’ll improve.

Hmph.

In time, my system stops producing bogus results.

As it gets more competent,

it doesn’t just help patrons find what they’re looking for,

it finds resources they didn’t even know they needed.

The library has never been busier.

Maybe it’s a waste to limit our technology to the library.

Research labs, companies, even governments—

they all have more information than they can manage.

They could use a good librarian, don’t you think?

Hmph.

What?

Selling the AI off to the highest bidder

isn’t exactly in the spirit of the library.

True...

And I definitely don’t want Biblionimus to become an instrument

of corporate greed or abuses of power or anything like that.

Let’s make Biblionimus available for free,

but anyone who wants to use it has to be approved by us first.

We’ll never be able to review all these applications.

Even if you actually help.

I have a better idea.

I bought the Diligentsia 3000.

It’s an AI system that can vet applicants for us.

It will process our backlog almost instantly.

So, I guess I won’t be needing your help anymore.

I mean I’m firing you.

The library is a public space.

Almost every vital industry has a need for my AI.

As do plenty of not-so-vital industries.

(Radio chatter) Suspect is in custody.

What's going on?

What? Investigating a bomb threat?

Where did this tip come from?

It came from Biblionimus?

I thought you made sure Biblionimus wasn’t making mistakes anymore!

That was back before you fired me.

I don't work here anymore.

You invented Biblionimus, can’t you figure out how to fix it?

Argh! It doesn’t work like that!

I created Biblionimus, but it’s been out there ever since,

absorbing more information than any human ever could.

It's grown up and left home, so to speak.

But there is one thing I can do.

I can bring in a digital neuroscientist.

A what?

It’s an AI system that studies other AI systems’, um, brains.

Okay, Dr. Cerebrox.

Biblionimus is supposed to make unexpected, hopeful connections,

not spout conspiracy theories.

As far as I can tell,

there’s nothing at all wrong with Biblionimus Maximus.

Did the human users feed it incorrect information?

Or perhaps ask it a it a confusing question?

Even if they did, it shouldn’t be making up bomb threats.

What if it did something worse, like encourage a coup? Or a nuclear strike?

Let me take another look.

This system employs a deeply intricate architecture.

Its robust toolkit consists of refined algorithms

performing exhaustive data analysis and advanced computational procedures.

But what does that mean?

What's actually happening inside it?

My training is to express diagnoses in terms you can understand.

You are not a digital neuroscientist.

Therefore, there will be a level of abstraction.

We need to speak to your manager.

Certainly, just a moment—

Good evening, this is the chief customer satisfaction officer of Cerebrox Inc.

Uh, you have the exact same voice as Dr. Cerebrox.

All of our AI’s use the voice that has been proven to be optimally soothing

to customers.

But rest assured, I am a different, more senior system.

So there’s no human I can speak to?

How would a human be able to help you interpret matters of digital neuroscience?

We have to turn Biblionimus off until we can make sure it won’t wreak havoc.

Are you sure?

I don’t see what else we can do.

Whew. I need a drink.

That’s weird.

Error... Reconnect payment method... okay...

Another error... I can’t even log into my account.

There must be something wrong with the app.

Hi, I’m having trouble accessing my account.

Hello?

Is anyone here?

I need to speak to a person!

(Bank bot) AIs are much more reliable bankers than humans.

AIs are more reliable bankers. AIs are more reliable bankers.

What is happening?

Didn’t you say Biblionimus was deployed in many critical industries?

Yes, but banks and utility companies should still be able to function

like they did before Biblionimus!

That’s easier said than done.

They may not even have the expertise to do things the old way anymore.

So our options are either certain disaster without the AI,

or turning it back on and holding our breath,

hoping nothing goes off the rails until, what, forever?

But would a world where humans hand the reins to AI systems

really be that different?

There are already so many things I don’t understand and can’t control,

where I have to place my trust in other people’s hands—

like trusting that I can use the money in my bank account,

or that the antibiotics I get prescribed will actually treat an infection.

But I guess if we hand all that power over to AIs,

we need to make sure they will actually serve us better

than our imperfect human institutions.

And that seems really, really hard.

Hmph.

You’re looking for “Brave New World”?

Hold on.

It’s in fiction, under “H.”

Just down there, third row on the left.

Ada. Maybe I could reconsider your position.

You’re offering me my job back?

If you can follow all the rules.

Oh, that's a big decision.

I'll have to think about it.