Shakespeare, Artificial Intelligence and the Danger of Stopping Thinking
Shakespeare had a point. Four hundred years early.
More than 400 years ago, William Shakespeare wrote a conversation that could easily belong in today's debate about artificial intelligence.
In Henry IV, Part 1, a Welsh nobleman called Owen Glendower boasts that he can summon spirits from the depths. Hotspur, a young soldier, is unimpressed.
He replies:
“Why, so can I, or so can any man;
But will they come when you do call for them?”
What a wonderful question.
Glendower claims extraordinary powers. Hotspur wants to know whether they actually work.
No lengthy explanation. No complicated theory. Just a simple request for evidence.
Shakespeare would have made an excellent auditor.
Although perhaps a rather unpopular one. Four centuries later, we are surrounded by extraordinary claims about artificial intelligence. And we could do worse than borrow Hotspur's question.
“BUT WILL THEY COME?” — Shakespeare Meets Artificial Intelligence
Hotspur studies a spectacular AI apparition with a sceptical eye.
The golden network represents the power and promise of AI; its dissolving edges suggest uncertainty.
The central idea: Something can look extraordinary without being entirely trustworthy.
The danger is not the machine. It is us.
Artificial intelligence is remarkable.
It can write, translate, analyse, calculate, generate images and help solve complicated problems.
It can complete in seconds tasks that once required hours.
I use it myself, and I am frequently amazed by what it can do.
But there is something that worries me more than the technology itself.
The danger is not that machines will suddenly become intelligent enough to replace human judgement. It is that we may become so impressed by their intelligence that we stop exercising our own.
That, I believe, is the real risk. Not simply that AI might get something wrong.
But that we might stop noticing. Or, worse, stop checking.
When a good answer becomes a dangerous one
Imagine asking an AI system to analyse a complicated contract.
Within seconds, it produces a summary.
The language is clear. The structure is excellent. The conclusions appear reasonable.
But what if it missed an important clause?
What if it misunderstood a legal obligation?
What if one of its conclusions was completely wrong?
Would you notice?
Perhaps. But perhaps not. And that is where the danger lies.
We tend to trust things that appear intelligent.
A confident speaker can persuade us even when the argument is weak.
A beautifully presented report can conceal poor analysis.
An impressive dashboard can make an organisation look safer than it really is.
AI takes this problem to another level.
It can produce convincing answers faster than we can verify them.
And the easier it becomes to obtain an answer, the more tempting it becomes to accept one.
We may save time producing information, only to lose the habit of questioning it.
When thinking becomes optional
Another example.
An internal auditor uses AI to analyse thousands of transactions.
The system identifies several risks and prepares a report.
The auditor reviews it, makes a few changes and sends it to management.
The whole process takes a fraction of the time it once required.
Wonderful.
But did the auditor understand the analysis?
Were the assumptions checked?
Was the evidence sufficient?
Could something important have been missed?
Or did the auditor simply approve a conclusion because it looked professional?
This is not an argument against AI. Quite the opposite.
AI offers enormous opportunities to improve audit quality, identify risks and free professionals from repetitive work. But only if the time saved is used wisely.
If technology saves us three hours of analysis, perhaps we should spend some of that time asking better questions. Not simply producing three more reports. Otherwise, we risk becoming faster without becoming better.
And efficiency without judgement can be a very expensive achievement.
The question we should never stop asking
This is where Shakespeare becomes particularly useful.
Hotspur does not accept Glendower's claim simply because it sounds impressive.
He asks for something more.
Evidence.
We should approach AI in much the same way.
When a system produces a conclusion, we should ask:
What evidence supports it?
What assumptions has it made?
What might it have missed?
How could we establish that it is wrong?
Would I reach the same conclusion without the machine?
The last question is especially important.
Not because we must always disagree with AI.
But because we should remain capable of independent judgement.
A machine can help us think.
It can challenge us.
It can offer perspectives we had not considered.
But the purpose of using AI should be to improve our thinking, not remove the need for it.
The best use of artificial intelligence is not to stop thinking. It is to think better.
Intelligence is not the same as wisdom
There is another distinction worth remembering.
Intelligence and wisdom are not necessarily the same thing.
A person can know a great deal and still make terrible decisions.
A machine can process extraordinary amounts of information without necessarily producing the right recommendation for a particular situation.
Making a good decision requires more than finding an answer.
It requires understanding the context.
Recognising uncertainty.
Considering the consequences.
Questioning assumptions.
And sometimes having the courage to disagree with an apparently convincing conclusion.
The ancient Greeks had a word for this kind of practical wisdom: phronesis.
It is also the inspiration behind Phronesium.
Because better decisions do not come simply from having more information.
They come from asking better questions and exercising better judgement.
AI may give us more answers than any generation before us could have imagined.
That does not mean we have automatically become wiser.
Back to Shakespeare
Let us return to Glendower and Hotspur.
One man claims he can summon spirits.
The other asks whether they will actually appear.
It is a simple exchange.
But it captures something essential about human intelligence.
The ability to question.
To doubt. To distinguish what sounds convincing from what deserves to be believed.
These are not weaknesses. They are strengths.
And we should be careful not to surrender them simply because a machine can answer more quickly.
Artificial intelligence will continue to improve. Some of its achievements will be extraordinary.
We should welcome them.
But the more capable these systems become, the more important our own judgement becomes.
Not less. There is also a small irony in all this.
This article was written with the help of artificial intelligence.
The machine helped organise the ideas, develop the argument and find the words.
It did a rather good job.
Or so it would have us believe.
But does that mean the argument is correct?
Does it mean every statement is accurate?
Does it mean you should accept the conclusions?
Of course not.
You should question them.
And so should I.
Because the greatest achievement of artificial intelligence should not be making human judgement unnecessary.
It should be making human judgement better.
Shakespeare understood the value of a good question.
Four hundred years later, surrounded by machines that seem to have an answer for almost everything, perhaps we should remember his lesson.
The question is not whether machines can think for us.
It is whether we will remember to think for ourselves.
But will they come?
PHRONESIUM
The first element of better decisions.
Better questions. Better judgement. Better decisions.
