August – Artificial Intelligence and Dublin WorldCon

Managing The Machine was the title of a talk by Dr Genevieve Bell from Australia at Trinity College. This highly interesting and entertaining lady told us how she studied Anthropology in Australia, then moved to Stanford in California to complete her PhD. Amusingly a chat with a guy in a bar led to her being phoned at home by Intel, who had rung around colleges asking for the contact details of a red-headed female Australian anthropologist.

What Intel needed was social sciences, their female staff member explained during a meeting. They had computer engineers and most of them were male. They wanted help with two issues: women, and the rest of the world. This was 1998 and Intel (who specialised in semi-conductor manufacturing) were trying to work out how to sell computers to women. Rest of World was an odd term so Genevieve queried it. World turned out to mean America, so RoW meant everywhere else “and you’re from there.”

At this point Genevieve was on college staff and had never thought of leaving education. But she realised that by working with Silicon Valley she had a unique chance to influence the future of society and help many more people. She worked with Intel for eighteen years so she now calls herself an anthropologist, technologist and futurist. She has returned to education as the Australian National University invited her back, saying it needed people who’d had outside careers for the Innovation Institute.

We are now in the fourth wave Industrial Revolution.
1 Mechanisation and steam power.
2 Mass production and assembly lines, electricity.
3 Computers and automation of many jobs.
4 Cyber physical systems and the Internet of Things.

The changes have occurred on different timescales in different countries. Often they accumulated and did not replace previous systems.
In France, within months of killing the King and decentring priesthood, the central education system changed from monasteries to polytechnics. The School of Engineering was established in 1794. The authority system had broken down so scientists were assembled and told that intellect must drive the nation forward. Switzerland and the US were next to set up similar science establishments. West Point’s Army Corps of Engineers was the first US engineering school, with all instructors from the Polytechnic in France. All engineers were apprentices. Australia needed engineers for mining in the 1850 – 60s. Turkey established the Sulieman School of Naval Engineering.

Mass production was about managing capital and this occurred from the 1880s. The University of Philadelphia developed a method of thinking about business money – Gross Domestic Product, Union talks and so on at their School of Business, at the request of business.

From 1946 to the 1960s computers were just faster calculators. By the mid 1960s, IBM and Honeywell were the big computer firms. Languages came along to programme computers - Fortran was developed in 1956 by Honeywell and Cobol in 1960. The US Government therefore depended on two computer firms and their technicians. Stanford worked for two years at developing a general computer language and created Computer Science. In December 1968 the Association for Computer Science in San Francisco handed out a ten page curriculum. This is still standard and every course teaches it.

Today we need to manage cyber-physical systems, the Internet of Things and Artificial intelligence. Autonomous vehicles and smart lifts in buildings, with no buttons, are part of this picture. We need a new applied science.
Www.3Ainstitute.org is seeking to develop this science.
Autonomy, Agency, Assurance.
This seeks to frame the right questions rather than to go problem solving.

1 Autonomy. Systems that don’t need reference to pre-written rules. But this word comes from philosophy and in Western culture is associated with Frankenstein’s monster, or the Terminator, rather than with safety and security. How do people live in this world? What do they like? Learner drivers are on the roads and are not so good as experienced drivers – autonomous cars would need to know. People dislike bots making phone calls.

2 Nature of agency. What are the limits of a robot or AI’s action? Who determines what it does and will a car crossing a border have to be updated for Brexit? Are the rules to be kept in the car or outside it? Can emergency vehicles override them? Do cars talk to traffic lights or are they on a separate system? We need also consider neo-colonialism as companies’ interests are not everyone’s.

3 Assurance. Safety and risk – are they left to the judgement of people or machines or both? The German government has a white paper that says the car must be safer for pedestrians than current ones but the maker is liable for damage. Roads would have to be well marked. All road users would use a common rule set. Meanwhile, machine learning users say they can’t really explain how an AI got to the answer.

4 Metrics. So far, we see data measuring gives an increase in efficiency and productivity. This does not account for environmental sustainability. Car companies will say it’s used for safety but others say it’s about data collection. “What you measure is what you make.” Computers are energy intensive, server farms now using 10% of the world’s power. What if energy efficient was a key metric? Native peoples may have different ideas about data collection to a foreign firm.

5 Relationships. Human /computer interactions are changing. An external interface like a keyboard or screen, may be something we wear in or on us. Human skills are enhanced, flying planes through bad weather for instance. We can also have nurturing robots. A robot could do a simulation of empathy to give social care and culture, for instance to the elderly. The Japanese have no expectation of empathy from robots but are fine with using them for elder care or child care. In Japan, Genevieve saw autonomous robots moving goods around and working; they were following tracks near the sidewalk, and a sign saying ‘Autonomous robots’ warned people to keep out of their way. The public don’t have a folk history of Frankenstein stories, so they are fine with this situation. “They’re being autonomous, yes, but it’s okay. The robots are our friends.”

A new discipline is to study this interaction. “She who asks the best questions gets the best conversations and can make a difference.” However, these are human-angled questions. Human priorities and object recognition are not on the computer decision making trees.

The twentieth century brought discourses on Capitalism and Democracy. While Capitalism then needed Democracy, it is clear in the twenty-first century that Capitalism no longer needs Democracy. Governments are dismantling and under great financial pressure from firms, so what is the trusted centre? Genevieve says we need to build a better future rather than try to roll back. Technology often reinforces inequities, but more people are living better lives than ever, and to date women have fought to improve life for women, and this needs to continue.

Other questions include what would it take to decolonise AI? Stories of western fiction would be removed. These are a western post-enlightenment construct. She also points out that using queer theory, objects are not one thing or another but can be both, using quantum entanglement. Pattern recognition however tends to be linear: A + B = C. We can allow supervised and unsupervised machine learning, but should insist on transparency of access to data. Cybernetics (how robots move and process problems) presently is being developed separately from AI around the world, including Chile, Russia and Japan.

Www.3Ainstitute.org based in the Australian National University, is seeking to develop this applied computer science of the future and would be interested in your questions.
My immediate question is, what do computers want? Would they ask to move out of Australia to somewhere colder? Would they ask for faster connections? What do you think they would ask?

Many thanks to Dr Genevieve Bell and the Trinity College Long Room Hub for this excellent talk.

My husband and I also attended an evening chat for volunteers who are helping with the 2019 Dublin WorldCon. The event was held in Hodges Figgis Bookshop on Dawson Street. We came away with posters and leaflets publicising the World Science Fiction Convention, which will be held in the Convention Centre on Dublin’s Docklands.
Memberships are being taken now and the price rises in increments as the date approaches. The membership will also enable you to receive the pack of materials nominated for the Hugo Awards so you can vote on the awards.

https://dublin2019.com/

And here is the just-released brilliant video showing you what to expect of an Irish WorldCon. Don’t be frightened, we promise to lock the dragons away before you arrive.

https://www.youtube.com/watch?v=XqEYR...

In keeping with my theme I am making an SF book free this month.
Dining Out Around The Solar System Part One. This a lengthy novella complete in itself, also contained within the full length book.

Donal, an Irish lad, and Myron, a Cockney-Jamaican mix, meet aged seventeen as book reviewers and trainee journalists with zine London's Eye. Before long they're onto the hottest stories in London.
In their future, Stansted Airport has been converted to a space shuttle base and Londoners are recruited to mine the asteroids.
While exploring the other planets, we found that they were all inhabited. Now those people are coming to Earth and looking for work. They're also opening ethnic restaurants in central London.
This first part of the story includes Donal and Myron's participation in the System Summit.

Free from August 30th to September 2nd.
https://amazon.com/Dining-Around-Sola...

https://www.amazon.co.uk/Dining-Aroun...

If you enjoy a book please leave a review, which helps other readers.

Visit my website www.clareobeara.ie for news, puzzles, books, reviews and events; also my blog on places with good disability access. I am adding book covers to Pinterest boards after I review the books, so feel free to find me on Pinterest.