i. Added Literature review
There are several resources available that simplify and popularise complex technologies, starting with Wikipedia. Other resources can be tremendously useful even with little to no technical background, such as semi-specialised press. See for example:
Academic papers are also an avenue to find information although the language might be less accessible without prior technical knowledge. Nonetheless, it’s worth searching on Google Scholar and other resources to find papers on the technology you are looking at, ideally in a similar context to yours or focusing on similar concerns.
Some NGOs with technical resources also publish materials than can guide you in understanding how a particular technology functions and how it can be used in specific contexts, for example:
To a certain extent, manufacturers’ websites can provide useful insights on the technology you are looking at and how it might function. Technical or promotional documentation for products these companies make can be great tools to understand the specifications of a given technology and provide insights on how it works. You may want to use methods such as Google Dorking to find companies’ official brochures and other documents that will help you in your quest.
As with any research, cross referencing what you find, and verifying with more than one source is key to avoid misinformation!
ii. Using and testing alternatives
Trying to find affordable equivalent systems and studying how these work can give you better insight into what goes on within the system you’re analysing. If you are looking at facial recognition systems for example, it can be useful to look for open-source projects that you can freely dissect such as this one.
Using these alternatives might require some technical knowledge and not be easily accessible to everyone. Tutorials and guides for beginners can get you a long way to setup and test these systems and should be considered as an easier way to approach this strategy. Similarly, some online course on “how to get started with X” can help you get a clearer understanding of how a technology functions. Introduction chapters of the D2L book on deep learning for example will help you understand the different elements at play in AI technology.
Organisations with technical expertise might also share guides, documentation, and methodologies to use systems they make use of in their work. For example, PI has a Data Interception Environment to analyse traffic from Android apps and has made it available to anyone.
It might be worth looking for other people who have done this sort of testing before, such as experts trying to find flaws or demonstrate bias in a given technology. Joy Buolamwini’s work on racist facial recognition systems is a good example of an expert testing a piece of technology to expose its weaknesses.
iii. Questioning experts
After you’ve done your research there might still be some questions left unanswered, some dots that you are not sure how to connect or simply things you don’t have the technical background to understand. Reaching out to experts in academia, specialised press or civil society organisations can be a helpful call in these cases.
When doing so we suggest explaining as clearly as possible what you are trying to do, for what purpose, what you have understood so far and ask questions as precisely as possible. Experts in the field will usually be less interested in giving a lesson about a given technology than helping you understand its application in a specific context. Writing down a list of precise questions you have with details about the context will maximise your chances of getting a response or a call with an expert.
In terms of who you should reach out to, you might want to start by looking at academics that have written papers on the technology you are looking at – in particular, if the focus of their research is one of the key risks you have identified. Alternatively, you can write to professionals working with the given technology as they will usually have a lot of hands-on experience on how it is being used. Looking for people in working groups, exchange groups and knowledge sharing groups is a good first step as it indicates a will to share and learn, increasing your chances of finding someone willing to help you. Asking on specialised online communities such as StackOverflow or Reddit can also get you very relevant information.
Some organisations such as PI also have technologists that you can try reaching out to. They might not have expertise on the technology you are looking at but could possibly point you in the direction of resources or other people you can talk to.
What our partners say:
Totally understanding the technology you are looking at is hard to achieve. ADC suggests accepting that you can’t necessarily know how all the involved blocks work and to try and get your work peer-reviewed to ensure that you are not saying something blatantly wrong. Making sure you have the basics right and that your analysis is based on verified information is more important than trying to understand everything and focusing on details you might have misinterpreted. With that in mind, ADC recommends limiting the scope of your work to a few things and focusing on them.