AI
Artificial Intelligence: The Mirror
We keep asking what artificial intelligence will become. Maybe the more important question is what we will become with it.
I just read Anthropic’s latest threat-intelligence report.
It’s disturbing.
Not because artificial intelligence has suddenly become evil.
Because it hasn’t.
We’re the ones bringing the intentions.
The report describes suspected state-linked hackers using AI for phishing, intrusion, malware development and data theft. Criminal groups using agents to scan enormous numbers of applications looking for vulnerabilities. Governments building surveillance systems. Influence operations creating fake identities and propaganda. Researchers working on autonomous weapons. Fraudsters operating thousands of artificial dating personas.
And then there are AI companies themselves allegedly attempting to extract the capabilities of more powerful models through industrial-scale distillation.
Different countries.
Different motives.
Different objectives.
Same tool.
And that’s what struck me.
For years we’ve been asking some version of the same question:
What happens when artificial intelligence becomes powerful enough to behave like us?
Maybe that’s the wrong question.
The more immediate problem isn’t that machines are becoming human.
It’s that machines are becoming extraordinarily capable while remaining available to humans who are already everything humans have always been:
Loving and hateful.
Generous and greedy.
Compassionate and cruel.
Curious and manipulative.
Creative and destructive.
The danger isn’t new, the leverage is:
AI didn’t invent the scammer.
It didn’t invent propaganda.
It didn’t invent surveillance.
It didn’t invent war.
It didn’t invent loneliness.
It didn’t invent exploitation.
It didn’t invent deception.
It didn’t invent tribalism.
It didn’t invent our seemingly endless appetite for power.
We did all of that ourselves.
AI simply introduces something humanity has rarely possessed at this scale:
Leverage.
One person can suddenly do the work of ten.
Then a hundred.
Perhaps eventually thousands.
Expertise that once required years of training can increasingly be accessed through a conversation. Tasks that once required teams of specialists can be delegated to agents that don’t sleep, get bored, demand salaries or question the objective they’ve been given.
That is extraordinary when the intention behind it is good.
Imagine what that leverage can mean for medicine, science, education, accessibility, engineering and human creativity.
But leverage doesn’t have a conscience.
Give a compassionate person leverage and they can help more people.
Give an entrepreneur leverage and they can build faster.
Give a scientist leverage and discoveries may happen sooner.
Give a criminal leverage and crime scales.
Give an authoritarian leverage and surveillance scales.
Give someone intent on killing people leverage and something considerably darker scales with it.
The technology magnifies the intention.
And perhaps that’s why these reports disturb us so deeply.
We want the danger to be the machine.
Because then we can imagine installing the right guardrails, writing the right laws, building the right kill switch and solving the problem.
But what if the machine isn’t the deepest problem?
What if it’s the mirror?
Human beings have always had extraordinary tools.
Fire warmed our homes and burned our enemies.
Chemistry created medicine and chemical weapons.
Nuclear physics gave us energy and Hiroshima.
The internet connected humanity and created an unprecedented infrastructure for deception, exploitation and manipulation.
The technology keeps changing.
The creature holding it doesn’t change nearly as quickly.
And now we’ve created something different.
A tool that doesn’t merely extend our muscles.
It extends our cognition.
Our ability to research.
To reason.
To persuade.
To code.
To imitate.
To strategize.
To deceive.
To create.
To destroy.
That’s an entirely different kind of leverage.
Which makes me wonder whether the great challenge of artificial intelligence ultimately won’t be technological at all.
It may be human.
Can our wisdom keep pace with our capability?
Can our moral development keep pace with our computational development?
Can people who suddenly possess extraordinary leverage also develop the restraint required not to abuse it?
I don’t know.
And perhaps nobody does.
But I know this:
AI doesn’t need hatred.
It doesn’t need greed.
It doesn’t need resentment.
It doesn’t need an ideology.
It doesn’t even need ambition.
We can supply all of those ourselves.
And we can supply something else too.
Compassion.
Curiosity.
Service.
Creativity.
Healing.
Love.
The same intelligence that can help someone discover a vulnerability can help another person defend against it.
The same technology capable of manipulating someone can help another person understand themselves.
The same extraordinary leverage can be directed toward relieving suffering instead of creating it.
So maybe artificial intelligence isn’t destined to save humanity.
And maybe it isn’t destined to destroy us either.
Maybe we’ve created something far more revealing.
A mirror with leverage.
And as its capabilities continue accelerating, the question staring back at us becomes harder to avoid.
Not:
What will artificial intelligence become?
But:
What will we become with it?