Artifical Intelligence, the Printing Press, and the Ottomans

The latest explosion of advances in generative AI are breathtaking, that’s for sure, but there is also growing concern that they may cause huge disruption and upheaval in social and economic life, from causing massive layoffs to the AI taking over control and even destroying human civilization. To prevent such catastrophy, some people are calling for a temporary pause on AI development, or even, a total, outright ban on AI.

I have come across an interesting counter argument against these calls for banning AI by Yann Lecun, who is the chief AI scientist at Meta. He gives the banning of the printing press in the Ottoman Empire until the 18th century as an example of what happens when you stand in the way of development. He argues that banning the printing press caused the Ottomans to miss the Renaissance, the scientific revolution and the Enlightenment, and fall behind Europe in the race for intellectual, scientific, technological and economic development.

Here are some instances where he makes this argument. He is quite active on twitter, and last year, he first referred to the Catholic Church attempting to ban printing, and then moved to the Ottoman Empire. A month later, he made the same argument, but this time, he cited an article on economic history, titled “The political economy of mass printing: Legitimacy and technological change in the Ottoman Empire”. In August, he argued against calls for banning AI in content creation. A month later, he tweeted again on the same topic, in reply to a tweet that in turn was a reply to another tweet by Lex Fridman, way back in 2019, who, posting a photo of an ad by a horse carriage maker against the disadvantages of automobiles, argued for the futility of resisting technological change.

In November 2023, we see Lecun supporting the UAE Minister for AI who regarded the Ottoman ban on printing press as the main reason for the decline of the Ottomans.

I also came across a couple of interviews he gave where he makes this argument, for example at CBS Morning, two months ago. And most recently, when he was interviewed by Lex Fridman.

Yann Lecun's argument is that even if "we" in the West ban AI research, the Chinese, the Russians, or any other entity (state or non-state) will continue developing their AI capabilities, and the West will fall behind, just like the Ottomans did. When the Ottomans realized their backwardness from the late 18th century onwards, it had already been too late to catch up with Europe. Similarly, Lecun warns, banning AI research and implementation might cause the West to fall behind others, which would then cause irreversible damage.

Stated as such, I agree with Yann Lecun that even if “we” banned AI or tried to introduce controls and ethical standards, “others” will certainly not do so. The genie is out of the bottle, the potential rewards are great, the promise of power is just so alluring.

Still, I think the argument “Banning the printing press cause the Ottoman decline” oversimplifies complex history and leads us to a familiar trap of misidentifying correlation as causation.

Firstly, the time scales involved in the printing press argument is enormous, covering two centuries or even more. The argument states printing press was banned in the Ottoman Empire in early 16th Century. And when did the decline of the Ottoman Empire become apparent, most importantly to the Ottomans? Towards the late 18th Century, or even early 19th Century. Many historical developments took place in this time period, from geographical discoveries to the Renaissance, Reformation and the Enlightenment in Europe, so much so that it becomes more and more difficult to identify the chain of cause-and-effect. When you eat a foul dish of chicken and you get sick the next day, it is easier to identify chicken as the cause of your sickness, and call it food poisoining. But when the effects of bad diet accumulate over many years and you become obese, get diabetes and all sorts of metabolic diseases, it becomes harder to point out a specific instance of eating a chocolate bar as the cause of all your metabolic problems.

The key phrase here is “cause and effect”. In the great flow of time and history, one event causes another and then the former is called the cause and the latter is caused effect. This creates an infinite chain of events that are linked together, one causing the other, and one becoming the effect of another. But it is often difficult to distinguish correlation with direct causality. Maybe both events were caused by an entirely different event? Hence, one must always look for the root cause. It is not possible to carry out controlled experiments in time, changing the parameters and then rewinding to an initial condition, to see how things play out differently.

In the case of the printing press, luckily, we have a historical example that may help us in this respect. In the 1730s, during the “Tulip Period”, the Ottoman sultan finally gave permission to Ibrahim Muteferrika, permission to establish a printing press until his death in 1747, he published about 17 books, each with a print run of about 500 copies. Overall, it is estimated that this first printing press published about 10 thousand book copies. Unfortunately, when Muteferrika died, about 3000 copies of books remained unsold. Although there were attempts to continue his work, they did not progress much, and the first printing press was closed shortly.

So, if banning the printing press in the 16th century was the reason for the Ottoman decline, allowing it in the 18th century did not cause much of a progress either. Books printed in the Muteferrika press were artfully produced but unfortunately they were not cheap. If there had been more demand for books, then maybe economies of scale could have been achieved, but alas, it did not happen.

It is important to underline the role of "demand" here. From the late 15th century onwards, there was massive demand for printed material in Europe, be it books or pamphlets. The demand came from the huge religious, social and political debate around the Protestant ideas about the position of religion and Church in individual, social and political life. This demand drove the establishment and then development of the printing press in Europe. In the Ottoman Empire, the publication of books, newspapers and magazines took off right after the 1908 Constitutional Revolution. It was a period of intense debate and the printing press provided the intellectual platform on which arguments were fought over.

Coming back to AI, there certainly is a lot of demand for more powerful, more capable AI models, tools and applications. I agree with calls for more responsible, ethical application of AI, but banning AI and hoping for a gentler development curve of AI capabilities will not simply work. It is already too late.