The dilemmas of privacy & surveillance

View comments Leave comment

Cyberspace must not be an unpoliced area of society – it is much too important for that. But the courts have ruled that mass surveillance of citizens by their Government is disproportionate and unacceptable in a democracy. After Edward Snowden revealed the existing extent of surveillance, Internet experts have strengthened encryption and the security services say that their ability to disrupt criminals have been weakened.

In the battle between the spooks and the geeks, who will win? And who should win?

Lecture date: Tuesday, 7th Feb 2017 – 6:00pm at The Museum of London, 150 London Wall, London, EC2Y 5HN

Available Downloads

Please feel free to download the lecture slides (powerpoint) and lecture transcripts (Ms Word)

Presentation transcript

Leave a reply

2 thoughts on “The dilemmas of privacy & surveillance

  1. Martyn Thomas CBE FREng says:

    Thanks for the comment, Jane, and apologies for not noticing it earlier! I agree entirely with your concerns – indeed, I raised related issues in this week’s lecture on Artificial Intelligence.

    When a data analysis system is based on machine learning (ML), it will inherit whatever prejudices exist in its training data (as the world saw with amusement when the Tay chatbot had to be taken off-line after it announced its support for Adolf Hitler and said it was smoking cannabis in front of the police).

    This stops being amusing when the ML system is taking decisions about medical treatment, loan applications, the cost of insurance, or whether to invite someone to a job interview, because most ML systems cannot explain the detailed basis on which the decision has been taken and even the programmers who designed and wrote the system cannot understand it in detail. The recent Royal Society report on ML is clear that these problems are still unsolved and require research.

  2. JN says:

    Hello Martyn, and thank you for another gripping lecture.
    One thing that’s often said in relation to surveillance is “If you’ve nothing to hide, you’ve nothing to fear”.
    I’d be interested in your comments on this statement.

    I have nothing to hide: but I certainly fear. And what I fear is mistakes.
    I’ve worked in Government Departments. They make mistakes.
    It certainly seems to me that with these vast stores of data, there’s plenty of scope for making mistakes in the pattern recognition. Or even not really mistakes, but let’s say “errors of judgement” due to series of co-incidences which lead to the wrong conclusion.
    For example: today I travelled on a bus – perhaps the person beside me on the seat was a known terrorist, so I might immediately be suspect. Then I spent some time standing in front of a government building. I was sketching, as it happens, but the software doesn’t know that. Then I went to a lecture, by Martyn Thomas, that well-known authority on computer surveillance, and I accessed his website. Put all that together and see the pattern: I’m clearly a terrorist, staking out a building, and then working out how to evade detection…..
    I’m not, but if they have the data, and I don’t, then it becomes difficult to defend oneself. What if this erroneous data were used in screening job applicants, immigration appeals, even ability to board an aircraft. One would be refused, and not even know why, just as now sometimes credit applications are refused based on data to which we have no access.

    So that’s why I think the balance is too far over on the side of surveillance. I’d be interested in your view.

Leave a Reply

Your email address will not be published.