Why Julian Assange shows that only intelligent machines can be truly rational

Is Julian Assange rational? There are no prizes for guessing the responses of most City A.M. readers to this question. He faces questioning over allegations (which he denies) in Sweden, and he claims that America will try to extradite him and put him in prison for a very long time. Assange has chosen instead to live within the confines of the Ecuadorian embassy in London for over three years.

On one level, economic theory can easily account for his behaviour. As a rational decision-maker, he estimated the potential costs and benefits of alternative courses of action, compared them to his preferences, and made what for him was the best possible decision. He probably considers the embassy to be preferable to life in a maximum security US jail. All standard economics textbook stuff: the costs of leaving the embassy outweigh the benefits.

But in forming his views on the costs and benefits, Assange had to make some quite difficult assessments of the likelihood of future events. What are the chances of him being charged by the Swedes? What is the probability of the Americans both trying and succeeding in getting him extradited? What is the likelihood of being found guilty if he is ever bundled over to the US? The answers are not obvious and there is no easy solution.

Even his ploy of getting the ludicrous United Nations panel to pronounce that he has been “arbitrarily detained” shows the inherent uncertainty around future events. Assange could have been almost certain that it would find in his favour. But it would have been hard for him to anticipate the ridicule with which this has been received, even among the liberal left. Making choices which involve future consequences is a much more difficult problem than simple economic choice theory might lead us to believe.

Assange could perhaps have benefited from the remarkable algorithm developed by a team at Google Deep Mind, and described in a paper in the top scientific journal Nature last month. Their work seems to be a very important step towards creating intelligent machines. The programme defeated the European Go champion 5-0. “So what?” you might say. But Go is an immensely harder problem than chess to attack by purely straightforward computational power. Success depends much more on a deep intuition of what does or does not constitute a good position. The astonishing feature of the algorithm is that it makes moves, in positions which have occurred previously in human games, which even the strongest human players simply cannot understand.

The implications spread far beyond the game of Go. The algorithm faced a challenging decision-making task. The precise consequences of a move cannot be computed. The optimal solution is so complex that it is infeasible to approximate. Yet the algorithm succeeded. Maybe humans have created the first truly rational thinker.

Paul Ormerod

As Published in City AM on Wednesday 10th January 2016

Image: Julian Assange by Global Panorama licensed under CC BY 2.0

Share this post



e: aobyrne@volterra.co.uk
t: +44 020 8878 6333

Alex O’Byrne, Associate at Volterra, is an experienced economic consultant specialising in economic, health and social impact, economic strategy, project appraisal and socio-economic planning matters.

Alex has led the socio-economic and health assessments of some of the most high profile developments across the UK, including Battersea Power Station, Olympia London, London Resort, MSG Sphere and Westfield. He has significant experience inputting to EIAs and s106 discussions as well as drafting economic statements, employment and skills strategies and affordable workspace strategies.

Alex is also experienced at economic appraisal for infrastructure. He was project manager of the economic appraisal for the City Centre to Mangere Light Rail in Auckland. He also led the economic and financial appraisals of the third tranche of the Transport Access Program for Transport for New South Wales, in which Alex developed and employed innovative methodological approaches to better capture benefits for individuals with reduced mobility.

He is interested in the limitations of current appraisal methodologies and ways of improving economic and health analysis to ensure it is accessible to as many people as possible. To this end, Alex recognises the importance of transparent and simple to understand analysis and ensuring all work is supported by a robust narrative.

Alex holds a BSc (Hons) in Economics from the University of Manchester and he was a member of the first cohort of the Mayor’s Infrastructure Young Professionals Panel.


Senior Partner

e: eevans@volterra.co.uk
t: +44 020 8878 6333

Ellie is a partner at Volterra, specialising in the economic impact of developments and proposals, and manages many of the company’s projects on economic impact, regeneration, transport and development.

With thirteen years experience at Volterra delivering high quality projects to clients across the public and private sector, Ellie has expertise in developing methods of estimating economic impact where complex issues exist with regards to deadweight, displacement and additionality.

Ellie has significant experience in estimating the economic impact across all types of property development including residential, leisure, office and mixed use schemes.

Project management of recent high profile schemes include the luxury hotel London Peninsula, Battersea Power Station and the Nova scheme at London Victoria. Ellie has also led studies across the country estimating the economic and regeneration impact of proposed transport investments, including studies on HS2 and Crossrail.

Ellie holds a degree in Mathematics and Economics from the University of Cambridge.