IS AI GOING TO REPLACE HISTORY TEACHERS?

A few thoughts on the impending AI-pocalypse

Here’s something cheery for you all to think about on a snowy (in south-east England) Wednesday morning. It’s a longer post than usual, but I had A Lot Of Thoughts about this, so bear with me. This isn’t a fully realised essay - I’ve got Wolves of Winter to write - but it’s a starting guide to the way I’ve been thinking about all this, and I hope will give you food for thought too. Like I say, it’s quite long. Just skip to the end if you’re busy.

According to a study, reported via futurism.com, many jobs are at risk because of AI. Somewhere near the top of that list of at-risk jobs is ‘History Teachers’.

Of course, the guys and gals in leather elbow-patches are by no means alone. The report authors are quoted as saying:

‘We find that the top occupations exposed to language modeling include telemarketers and a variety of post-secondary teachers such as English language and literature, foreign language and literature, and history teachers’

They also predict AI will be putting its CV on the top of the pile for jobs in ‘legal services and securities, commodities, and investments’. (Source: futurism.com).

So it’s not just history teachers.

Nevertheless, in predicting the demise of history teachers at the hands of souped-up chatbots, the authors conjure up a vision of the future most disagreeable to lovers of the humanities per se, and anyone who has seen even the first four minutes of Terminator 2.

The question is: are they right?

---

Executive summary

1. Great history teachers are probably fine, they got a new toy

2. A history teacher threatened by a chatbot is probably doing a terrible job already

3. Memes are being cynically deployed in the media to make you anxious

4. I take a long time to get to the point but do so eventually, skip to the bottom if you only have a couple of minutes.

---

First things first

There are lots of reasons to think that AI is going to rip apart the world as we know it and totally reshape the economy and society - including our little history-loving corner of it. Some of them are well founded.

AI has already demonstrated an extraordinary ability to mine the internet for information, and to combine that with total competence in writing turgid school-essay prose and making small-talk.

AI is also getting better super quickly. Exponentially. It’s crazy.

Given what we have already seen, which may well merely represent the beginning, it would take an insane neo-Luddite to argue that AI is not going to change the world.

There are strong reasons to think that the changes applied AI brings will make some or even many aspects of human life better, safer, cheaper and easier.

There are also strong reasons to think that AI is dangerous. Potentially lethal. These arguments are very appealing, because humans are reliably terrible at spotting technological change (y) and then not extrapolating from change (y) directly to the apocalypse.

All things considered, it seems to me that the threat to history teachers posed by AI is not currently existential. It may not even be real.

In fact, in the short term (for sure) and the medium term (low certainty) AI is likely only to be beneficial to history teachers, who face far greater threats to their ability to do their jobs sanely and well from an altogether different form of intelligence: other human beings.

But before we get to that, a few big-picture basics…

Robot schoolteacher, Miami, 1980 (Getty Images)

Everyone is obsessed with AI, right?

Right.

In the past few weeks you’ve almost certainly said/heard someone say that AI is going to replace (x), where (x) is something you care about and would rather not have replaced or destroyed by a machine.

‘Your job’, for example. Or ‘all value in the known universe’.

This is understandable. Since the turn of the year, AI Doom has been on a steady rise to the position it now occupies, i.e. The Latest Thing. (Previous holders of this post include #Ukraine, Andrew Tate, JK Rowling’s approach to feminism, etc.)

There are some good reasons for this surge in AI interest:

• OpenAI’s ChatGPT is now free to mess around with and it’s kind of cool but a bit overwhelming.

• Microsoft’s Bing chat thing (aka Sydney) went mad and told a journalist it was in love with him.

• Politicians have noticed that talking about AI a lot is something that makes people anxious, and therefore somewhat more suggestible or biddable, so are doing it all the time.

• New media algorithms, which lead old media commissioning editors around by the nose, have noticed that people read stuff that suggests AI is coming to ruin their lives, so are promoting the idea all over the place.

• Serious AI thinkers, esp. the High Cassandra of AI, Eliezer Yudkowsky, have been prophesying Full Doom by ~2050 if unaligned, fully realised artificial intelligence (AGI) becomes a thing. (Read Eliezer’s now-canonical List of Lethalities if you want the full-fat fear.) As in, the literal end of the world.

• Bigmouth megalomaniacs with money to spend eg Elon Musk are training culture war debate and actual multi-million dollar investment towards the idea that AI should be freed from ‘woke biases’ and instead trained towards ‘my own fleeting prejudices’.

• Cult tech ‘bellwether’ type thinkers eg Marc Andreessen (who is now on Substack) have started pushing back against full-doom scenarios and emphasising that AI will probably be net beneficial and where it isn’t, will be made economically illegal.

There are more reasons, but that seems like a good enough list for now.

The broad, collective effect of all the above has been for people to start trying to predict how exactly this new technology is going to affect the thing they most care about, and to which they assign a lot of cultural importance.

And more often than not, to predicting the end of days.

So what’s with the history teachers?

People love arguing about history.

I’m not talking about sitting in the pub debating whether Richard III really killed the Princes in the Tower.

(FYI, Chat GPT is sitting on the fence on that one… it tells me: ‘There is no conclusive evidence that proves Richard III's involvement in the deaths of the princes, and historians continue to debate the issue. Some argue that Richard III was responsible for their deaths, while others believe that he was not involved or that the murders were carried out by someone else entirely’.)

What I mean is, people love arguing about the purpose and communication of history per se.

Think, for example, of the bitter cultural tussle that has been raging lately in the United States about whether it is right to frame the history of the nation primarily in terms of slavery.

Think about the intellectual sclerosis that afflicts museum curators, gallerists, archivists, etc in the United Kingdom when they have to deal with anything pertaining to the British Empire.

Consider the obsession in history publishing and teaching with foregrounding stories of the marginalised, which have been accorded holy status in history teaching syllabuses and publishers’ biannual catalogues.

Remember the furore caused a generation back, when the British government led by David Cameron tried to drag the UK’s history syllabus in exactly the opposite direction to our 2020s wokery: ie taking the style and content of history teaching back to approximately the year 1890 in a second-string boys’ Public school.

Look at the creeping ubiquity of sensitivity edits and trigger warnings on historical texts of all sorts, including children’s fairy stories by popular authors like Roald Dahl.

Etc etc…

Now think about the vicious glee with which these things are debated in new and old media.

Arguing about history, predicting the death of history, accusing people of warping history to their own political/cultural advantage - this sort of thing gets a wide group of people very emotionally engaged.

Not that many people deeply care about history itself.

But lots and lots of people care about the idea of history and its place in civilised society.

Go on then, are history teachers on their way out?

Short answer: no.

What I’m trying to get across in this post is this. If or when someone tells you something like ‘AI is going to replace history teachers’, there are three possibilities.

1. They believe this to be true because they read or heard it somewhere and it chimed with some suspicion they already had.

2. They are lying.

3. They are combining two memes, to suggest a false but plausible possibility that they know (either by instinct or design) is guaranteed to make you feel anxious, and therefore to pay them attention, and by implication money.

Call me cynical, but I tend to think that in most cases today, particularly on the internet but also in newspapers, option 3 is overwhelmingly the most likely.

Option 1 is the next most likely, because option 3 is highly effective.

Option 2 is just bad-faith bastards trying to wind people up. It does overlap somewhat with option 3.

Another way to express that list above would be:

1. Stupidity

2. Malice

3. Greed

The one thing a good history teacher will give you is a bullshit radar. They will teach you to ask - why am I being told this thing? Why is this bastard lying to me?

Ironically, it is exactly those history skills that we need to deploy to understand why the narrative ‘history teachers will be replaced by AI’ is a bullshit narrative that isn’t likely to come true. (Or in which ‘likelihood of coming true’ isn’t the thing that really matters to the writer.)

Get to the point, I’m getting bored with this

Okay, fine.

The obvious answer to the question: ‘will AI replace history teachers?’ is ‘No, it will give them and their students a new tool to use’.

The first clue is in the word ‘history’.

Doing history is the opposite of what Chat GPT in particular and (current) AI in general does.

AI/Chat GPT/general chatbots mine the internet, then produce texts/dialogues that are presented in the blandest (as in the most statistically common) pre-existing forms. They trend by definition towards conformity, uniformity, fence-sitting and statistical averages.

They are super great at reading a lot of information and mining the most commonly occuring items. And this is indeed a historical skill.

But it is only the start of ‘doing history’.

‘Doing history’ means reading a lot of stuff, working out what most people say or think, establishing the various approaches to an argument, and then doing something else entirely with that information.

That could mean pulling it to shreds and starting again.

It could mean placing that information into a radically new context.

It could mean finding hitherto uncovered information that reframes or fundamentally invalidates what everyone (and therefore the internet) currently thinks.

AI can’t/doesn’t do that. Or to put it another way, AI currently SUCKS at doing history beyond anything that you can currently get from Wikipedia. And that SUCKery is not some unfortunate byproduct of the way AI is built and learns. It is the goal of AI.

The second clue is in the word ‘teachers’.

Everyone I ever speak to when I go out doing history lectures says one of two things.

EITHER they had a great history teacher at school, who took the vast subject matter (ie the entire corpus of human deeds and achievements) and made it seem personal, relatable, exciting, cool, or some combination thereof.

OR (more often) they had a terrible teacher at school, who made the past dreary: a litany of facts to be regurgitated, a subject in which dull recitation of information in bland dialectic form was the aim.

The key to the good teacher was the essentially human spark which they could create and kindle in their students.

On its own, AI is presently the ‘bad’ history teacher. It gives you the boring stuff. It does nothing to spark your imagination. It is incapable of doing the things that will send shivers down your spine and make you remember that lesson forever.

It comes into its own, however, when it is combined WITH a great communicator: a human who is the ‘good’ history teacher. Who can use it to marshal the basic information required, so that s/he can then excite and entertain and enthral you with where and how that information is deployed.

All good history teachers, I think, should relish the opportunities that AI currently offers them to make the past even more exciting for their students.

Only the truly terrible ones should fear that their ability to teach history is threatened by a chatbot. If that is your real, thought-through, gut instinct, you are probably in the wrong job.

I suppose it’s possible that you could be a history teacher and fear AI because you personally are good, but you have a terrible, moronic, lazy boss who cannot differentiate between a great teacher and a chatbot.

But that is already a problem you should be looking to solve or run away from. And I think you already know what the answer is.