Artificial Intelligence & The Bystander
More or Less Intelligent?
I have written one post about artificial intelligence (AI) every day the last 60 days and it has ranged from long discussions of new techniques within machine learning to fairness, policy, strategies, AI safety and much more.
One can hardly say I have been keeping it real? Rather one could say I have been keeping it artificial! Yet then again I have discovered that AI is far more than buzz. The field has real-life applications saving lives: whether it is elderly care, cancer or wildlife —the field of AI has a vital contribution that is so bombastic it cannot be ignored.
On the other hand these applications are often being made by a narrow group of the population. Adverse actions can occur due to unexpected or expected circumstances. When an application is rolled out to millions or even billions there is often a serious lack of diverse fields of backgrounds and a high degree of engineering as well as programming.
My assumption were that these teams would be less ignorant than I expected and I was right. Particularly writing about OpenAI and the research being undertaken in AI Safety has made me realise that there is a thorough effort by certain actors to behave responsibly.
This is increasingly important due to growing disagreement between people on this planet, particularly hegemonic or emerging powers racing to develop the most advanced computing technologies.
My concern is as much for the present as it is for the diverse future predictions. Machine learning is often a set of predictions that shape new predictions; this is developing algorithms, the field of AI. The only certainty is that I need to learn more.
Learning more means attempting to grapple with eclectic proportions of a massive field of knowledge within the social sciences. If I count the pages I have read this year and the words it would be considered big data. Beside this I need to improve my skills in Python and R, being technical, while coming to terms with my hunger for more mathematics.
Saying this makes me sound like a protagonist, yet I feel more like a bystander in a protest. I was there in Oslo on the 15th of March to protest the climate crisis, yet I did not shout and hold a banner. I talked to youth to try to understand what they wanted to change and how.
In the same way when I talk to developers working in the field of AI that are so deep into their work I feel like a bystander. As if I am looking just under the surface seeing the unfathomable deepness and dark blue beyond my understanding.
Then again these developers does at time seem so deep into their work that they disconnect with humanity or forget the people they serve and design days that revolve around committing code to online repositories.
Thus perhaps I am not a bystander, rather I am participating. The field of social sciences where my focus lies is on participatory observation also known as the ethnographic method.
Therefore in participating it may mean committing text and committing code. As much as this is the case it means seing how universal or partly customised assumptions play out in real-time or for those who are on the receiving end. Falsification in my belief is of the uttermost importance to make progress with equal amounts of care and concern.
It seems we are ever closer to understanding intelligence, yet are we more or less intelligent? In fact this does not matter as much, although it is important.
Intelligence as described in dictionary is: (1) the ability to acquire and apply knowledge and skills; or (2) the collection of information of military or political value. This is surely not enough!
The ability to understand something; comprehension; sympathetic awareness or tolerance. This is surely partly what we are lacking in the current machine behaviour.
In this paradoxical manner I am critical and very hopeful. I am a bystander or observer, and I am certainly participating.
This is day 61 of #500daysofAI