Home » Accountability » – Ethics of Eavesdropping

Want yet more email?

Enter your email address to get notifications of new posts.

Request contact
If you need to get in touch
Enter your email and click the button

– Ethics of Eavesdropping

It has been recently reported (e.g. see: Bloomberg News ) that the likes of Amazon, Google and Apple employ people to listen to sample recordings made by the Amazon Echo, Google Home and Siri, respectively. They do this to improve the speech recognition capabilities of these devices.

Ethical Issues

What are the ethical issues here? The problem is not with these companies using people to assist in the training of machine-learning algorithms in order to improve the capabilities of the devices. However there are issues with the following:

  • While information like names and addresses may not accompany the speech clips being listened to, it seems quite possible that other identification would potentially enable tracing back to this information. This seems unnecessary for the purpose of training the speech recognition algorithms.

  • It has been reported that employees performing this function in some companies, have been required to sign agreements that they will not disclose what they are doing. To my mind this seems wrong. If the function is necessary and innocent then companies should be open about it.

  • These companies do not always make it clear to purchasers of devices that they may be recorded, and listened to, by people. This should be clear to users in all advertising and documentation.

  • The most contentious ethical issue is what to do if any employee of one of these companies hears a crime being committed or planned. Another situation arises if an employee overhears something that is clearly private, like bank details, or information that, although legal, could be used to blackmail. In the first situation, are these companies to be regarded as having the same status as a priest in a confessional or any other person that might hear sensitive information? A possible approach is that whatever law applies to human individuals, should also apply to the employees and the companies like Amazon, Google and Apple. So in the UK for example, some workers (such as social workers and teachers) who are likely to occasionally hear sensitive information relating to potential harm to minors, are required to report it. In the second case, companies could be legally liable for losses arising from the information being revealed or used against the user.

It seems likely that companies are reluctant to admit publicly that interactions with these devices may be listened to by people, is because it might affect sales. That’s does not seem a good enough reason.

Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

How do we embed ethical self-regulation into Artificial Intelligent Systems (AISs)? One answer is to design architectures for AISs that are based on ‘the Human Operating System’ (HOS).

Theory of Knowledge

A computer program, or machine learning algorithm, may be excellent at what it does, even super-human, but it knows almost nothing about the world outside its narrow silo of capability. It will have little or no capacity to reflect upon what it knows or the boundaries of its applicability. This ‘meta-knowledge’ may be in the heads of their designers but even the most successful AI systems today can do little more than what they are designed to do.

Any sophisticated artificial intelligence, if it is to apply ethical principles appropriately, will need to be based on a far more elaborate theory of knowledge (epistemology).

The epistemological view taken in this blog is eclectic, constructivist and pragmatic. It attempts to identify how people acquire and use knowledge to act with the broadly based intelligence that current artificial intelligence systems lack.

As we interact with the world, we each individually experience patterns, receive feedback, make distinctions, learn to reflect, and make and test hypotheses. The distinctions we make become the default constructs through which we interpret the world and the labels we use to analyse, describe, reason about and communicate. Our beliefs are propositions expressed in terms of these learned distinctions and are validated via a variety of mechanisms, that themselves develop over time and can change in response to circumstances.

Reconciling Contradictions

We are confronted with a constant stream of contradictions between ‘evidence’ obtained from different sources – from our senses, from other people, our feelings, our reasoning and so on. These surprise us as they conflict with default interpretations. When the contradictions matter, (e.g. when they are glaringly obvious, interfere with our intent, or create dilemmas with respect to some decision), we are motivated to achieve consistency. This we call ‘making sense of the world’, ‘seeking meaning’ or ‘agreeing’ (in the case of establishing consistency with others). We use many different mechanisms for dealing with inconsistencies – including testing hypotheses, reasoning, intuition and emotion, ignoring and denying.

Belief Systems

In our own reflections and in interactions with others, we are constantly constructing mini-belief systems (i.e. stories that help orientate, predict and explain to ourselves and others). These mini-belief systems are shaped and modulated by our values (i.e. beliefs about what is good and bad) and are generally constructed as mechanisms for achieving our current intentions and future intentions. These in turn affect how we act on the world.

Human Operating System

Understanding how we form expectations; identify anomalies between expectations and current interpretations; generate, prioritise and generally manage intentions; create models to predict and evaluate the consequences of actions; manage attention and other limited cognitive resources; and integrate knowledge from intuition, reason, emotion, imagination and other people is the subject matter of the human operating system.  This goes well beyond the current paradigms  of machine learning and takes us on a path to the seamless integration of human and artificial intelligence.

%d bloggers like this: