In Sapiens, Mr. Harari makes the argument that the ongoing development of humans, leading to the creation of AI technology will bring about a divergence between intelligence and consciousness (conscience?). This is also the idea elaborated on in more ominous terms by Mr. Bostrom in his book Superintelligence.
Ruminating on the above got me thinking about the ideas described by John Erskine in the early nineteen hundreds “The Moral Obligation To Be Intelligent”. It is a short essay in 5 delightful chunks. With World War I horribly close, he made one of the most civilized arguments I have read – intelligence is a virtue we are responsible to develop in ourselves.
I subscribe to his views and believe that our collective conscience is directly linked to the development of the faculty we call intelligence. For this to be the case however, we will need to address an issue which may be a lot more problematic now than it was in Mr. Erskine’s time. We need to radically expand our notion of intelligence from simple numeric characterizations (for example, IQ metric), to a more nuanced multi-dimensional description, of the likes offered by Howard Gardner.
Standardized test driven ideas of intelligence are perhaps damaging our societies and culture more pervasively than AI. Prevailing models of intelligence are also to blame for the lack of imagination in teaching and learning – in both human and machine varieties. The tendency to discretize and quantify what is to be learnt about ourselves and our relationship with the world is more harmful than we can imagine. This is certainly a threat.
The etymology of the word “Conscience” as described in an essay by Ms. Korsgaard, a Professor of philosophy at Harvard, hints at the far from adversarial relationship between intelligence/knowledge and morality:
… Latin “conscientia,” a direct translation of the Greek “syneidesis.” This ranges in meaning from being aware of something (hence our “consciousness”) to “knowing something in common with” someone
She concludes the essay with a point, referring to Nietzsche and Freud, that I think is the other pole in the argument between intelligence and morality that I started this post with:
… the vindication of our moral nature depends on whether the self-mastery which we get from conscience can be detached from social and parental authority, and exercised instead in the name of a set of sane and realistic standards derived from reason
We could end the topic where she does and we are already less confident that intelligence is opposed to morals. However, rather than rely on ill defined terms such as ‘sane’ and ‘realistic’ we should go further.
There is an urgent need today for Mr. Erskine’s criticism of the still-live-and-kicking Anglo-Saxon suspicion of intelligence in favor of morals. He reminds us of the ancient Greek tradition – where intelligence and morality were synergistically responsible for the civilized growth of a human. While he focuses on the Western tradition, I would like to add that his general theme is not divergent from the emphasis on concentrated, contemplative and dispassionate study one finds in other Eastern traditions, say in the Buddhist texts. The ancient Indians, also, lauding ritualistic but comparable forms of studious practice, called their oldest texts Veda – Knowledge.
I think the suspicion of artificial intelligence being popularized by some loud voices is misguided. We should instead be remediating and expanding our notion of human intelligence. The machines we create will follow consequentially. We can then fearlessly encourage the development of intelligence as enthusiastically as possible.
Reminding us of the arguments made by John Erskine in 1915, this is my contribution to the Human Vs AI debate – .
- Intelligence (of the Gardner variety) connects our experience to those of others, the environment and history.
- Taken to its limits, by artificial or natural means, will lead to Morality – allowing us to be con scientia (with knowledge).
Intelligence (of all kinds) is Good.