Posted in

Data, Metadata, Algorithms & Ethics

by Andrew Heikkila•
Intel Data Center GPU codenamed Crescent Island architectural slide showcasing Xe3P AI optimized GPU IP, up to 480GB LPDDR5x memory capacity, and a 350W air-cooled PCIe form factor.
Intel’s Crescent Island GPU targets AI inference economics by prioritizing massive LPDDR5x memory capacity over costly HBM architectures within a practical 350W air-cooled design.

The topic of ethical big data use is one that will likely continue popping up in the headlines with increasing frequency in the coming years. As the IoT, AI, and other data-driven technologies become further integrated with our social identities, the more discussion regarding its regulation we will see.

Recently, transparency advocates began pushing The Open, Public, Electronic and Necessary (or OPEN) Government Data Act, which aims to publish all non-federally restricted data in an open source format, allowing for standardized use by the government as well as the public.

Our federal government is not just the largest organization in human history, it’s also the most complex, said executive director of the Data Coalition, Hudson Hollister, in an article on the Federal Times. To conduct oversight across such scale and complexity is a daunting challenge, fortunately, that is where transparency comes in. By giving Americans direct access to their government’s information, we can deputize millions of citizen inspectors general to help this committee fulfill its mission.

This type of standardization, transparency, and ethical foresight aims to create a fair and balanced framework for the use of Big Data. Considering the pace of automation and IoT growth, these standards could begin affecting every industry under the sun, and perhaps ethical regulation is now more necessary than people realize.

Ethical Data Regulation

Big Data ethics became a very hot topic in 2013, when details of the NSA‘s PRISM program were made public by security contractor Edward Snowden, revealing that the government was collecting both data and metadata of its civilians without warrants.

While this big event may have sparked the conversation, it’s all but been resolved. More recently, there’s been significant disagreement on the right course of action to take when metadata mistakes are made, specifically concerning data mining and notification duties, according to legal expert Keith Swisher. Client information must be protected by their lawyers at all costs–but this has been complicated by the abundance of metadata attached to modern electronic documents.

Christian Fuchs writing for Huffington Post presents another recent, more extreme situation, in which YouTube lost advertising partners after these partners realized that some of their ads had been run alongside videos featuring anti-Semitism, white supremacist ideology, hate speech inciting violence, and right-wing extremism. Fuchs continues:

The commodity and the algorithm are blind for morality and meaning. They empty online communication from its meaning and fill the created void by automatically presenting advertisements. The algorithm does not care if the content it matches with ads contains dog poo, beef stew or a fascist cuckoo. The logic of algorithms and the logic of the commodity know no ethics and no morals. They only know the instrumental logic of numbers. The immorality apparent in the pairing of fascist videos with mainstream organisation’s ads is the very consequence of the automation of human activity in big data capitalism.

These three examples cover just about everything inherent in the Big Data ethical dilemma: data, metadata, and algorithms/data use.

Data, Metadata, and Algorithms

Everybody knows what data is. Since the dawn of written history, man has been creating data–but only in the last few years has there been the massive amount of data produced needed to fuel big data. In fact, there was more data produced last year than there has been in all of human history.

This is where metadata comes in. For those who are unsure of what metadata is, it’s essentially data used to describe larger sets data. For example, search engines analyze website meta tags to discern better what the larger data set (actual content on the website) is about. A more real world example might be phone metadata. This isn’t the actual content of your call, but rather who you called and at what time. As Jay Stanley and Ben Wizner put it in their article for Reuters:

Repeated calls to Alcoholics Anonymous, hotlines for gay teens, abortion clinics or a gambling bookie may tell you all you need to know about a person’s problems. If a politician were revealed to have repeatedly called a phone sex hotline after 2:00 a.m., no one would need to know what was said on the call before drawing conclusions.

Algorithm and data use is explored in the above quote–but it should be mentioned that this is just the tip of the ethical iceberg. The way that we utilize these resources now may serve as precursors to the way that we use them later. Ian Sharp, speaking at Big Data World in London, mentions that the next step is recognizing emotion in audio and video, which could eventually lead to the rise of AI.

The thing we all talk about is the rise of cognitive services, deep learning and potential use of robots ¦ [Microsoft CEO] Satya Nadella talks about democratising AI and [Oracle founder] Larry Ellison says every technology since fire has been used for good and bad, he said. It’s about finding the right, ethical user cases.

ICO’s Big Data Use Guidance

Big Data isn’t going anywhere. Healthcare, manufacturing, insurance, and driver safety promise to be revolutionized by the power of data and analytics, and these benefits will only continue to compound. This means that ethical guidelines will have to be established sooner rather than later. Fortunately, the Information Commissioner’s Office (ICO) has issued their own guidance on Big data, artificial intelligence, machine learning, and data protection. Their six key recommendations are:

  1. Using appropriate techniques to anonymize personal data.

  2. Providing meaningful privacy notices so organizations are being transparent about the processing of personal data.

  3. Embedding a DPIA framework into big data processing activities;

  4. Adopting a privacy-by-design approach in the development and application of big data analytics.

  5. Developing ethical principles to help reinforce data protection principles.

  6. Implementing innovative techniques to develop audible machine learning algorithms.

By following the proper guidelines and establishing the right regulation, Big Data can work for the people at large, and not just the companies wielding the data.

I'm a Big Data, IoT nerd who is also a performing artist out of Boise, ID. I started working in IT while I was attending College of Idaho '08 to '12 and then moved into web development and social/internet marketing and blogging shortly after. After ghost-writing a couple of whitepapers on data warehouse management software, I slowly but surely found myself increasingly interested in Big Data and Analytics and how it's seeping into basically every aspect of our lives. This opens up a whole new world of possibilities--both good and bad. I'm here to write about them. Follow me on Twitter @AndyO_TheHammer

Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.