Why Care About Artificial Intelligence Policy?
A Variety of Reasons to Engage with Policy
When you think about the word ‘clarity’ a policy paper may not be likely to be the first thing that comes to your mind. It is fair to say that it may not come to cross your mind at all, at any conceivable point, because why should you care? There is enough to do simply building complexity, or when navigating the various desires of clients or stakeholders in a project. After hours of hard work and the job is done at the end of the day I would imagine only a select few rush home with glee to read policy papers.
Luckily for you I am one of those, and I will attempt to tell you about excitement and why you might find it fun too. Although I feel slightly like someone expressing the joy of doing your laundry or a menial task, as reading policy papers seem to garner this order of response amongst professional developers.
Today I had the pleasure of speaking to a room of developers in Norway and asking the question: have anyone read the national AI strategy? The answer was as expected ‘no’ from most in the room with the exception of the economist that had invited me to speak.
The first point is that there is a burst of activity globally in regards to artificial intelligence and policy.
Although states have lost some power or ceded some ground to private companies they still facilitate to a large extent several areas of your life.
I began my talk with the extent of the current AI strategies around the world, and that states are getting actively involved to facilitate artificial intelligence.
Each policy is however different and has nuances that may help to understand when implementing projects. The US strategy is more defence-oriented. The Chinese one takes sustainability into account to a larger extent. France has the most climate friendly AI strategy with an understanding of circular economy. Europe is concerned with trustworthy and ethical AI as a competitive advantage etc.
Then I proceeded to show how Europe is doing by some measurements.
Showing as well the difference in collaboration within different countries — there is a clear trend in the most successful countries to have more collaboration than elsewhere.
There is a great increase in patenting particularly in America.
After showing a few graphs from the AI Index by Stanford University I showed a picture of the Tortoise AI Index.
Because my presentation was in Norway and the Norwegians often compare themselves to other Nordic countries I extracted the figures from those who had been on the list from their corresponding placement. Explaining that Finland was the frontrunner in AI within this group, particularly due to its great infrastructure.
I did brag slightly about meeting the minister of digitalisation and taking part in two of the participatory meetings for the development of the artificial intelligence strategy, and telling that the report was out the 14th of January, earlier the same month.
I attempted to explain that the Future of Life Institute is helpful in navigating AI policies in different countries, and that I thought it was wise to contribute to this effort of making it more accessible.
To help the developers consider what elements of the Norwegian strategy they could approach I made a list of ten points.
Then I proceeded to explain why the Norwegian case is different at least in terms of funding ICT research. The general trend in Europe with the numbers given previously stated that the government was the largest funder of research, while these numbers indicate that businesses provide the largest portion of investment for research in Norway.
However the question still remains.
So I decided to discuss this with the developers. For the developers it seemed like the understanding of new auditing from the state such as the GDPR regulations would be of interest. How would the state audit artificial intelligence?
Why do you build AI applications? If they are meant to be useful then they will likely be connected to people in some way, and in general people are citizens as well as ‘users’. Great governance of technology can be an amazing way to enrich people’s lives, while horrible governance of technology can at its very worst kill people.
Developing and deploying responsible AI applications can mean understanding government priorities.
This is #500daysofAI and you are reading article 233. I am writing one new article about or related to artificial intelligence every day for 500 days. My current focus for 100 days 200–300 is national and international strategies for artificial intelligence.