Every time you search Google, examine your Facebook feed, book a ticket, extract money from an ATM, or use a GPS, you are interacting with an algorithm.
A powerful recipe: sets of different codes combine to form algorithms, says cyber security expert Dr Muhammad Ikram.
An algorithm is a set of coded instructions designed to perform a specific task. At its most simple, it is like a recipe. It needs process. Making an omelette uses an algorithm: you take a frypan, an egg, whisk the egg, put oil in the pan, warm it, add the egg, cook it, flip it, plate it up. You might add spices, maybe a few drops of milk.
Algorithms have become our best friends: there’s Google page rank, changing the way we search for useful content; Uber for its seamless connection between order and delivery or coming and going; and the various forms of flight scheduling and booking which have revolutionised travel.
Online services use various combinations of simple algorithms (sorting and searching) and advanced algorithms (optimising the shortest path between two connected points after mining and sorting through millions of data points).
Advanced algorithms are transforming industries such as health care, education and space exploration.
Algorithms are truly an artform. The industry’s trailblazers are proof that having multi-disciplinary thinking helps.
British mathematician Alan Turing was a theoretical computer scientist who formalised the concepts of algorithm and computation. British-Canadian cognitive psychologist and computer scientist Geoffrey Hinton is regarded as 'the godfather of deep learning' (machine learning methods based on artificial neural networks). And Stanford cyber security innovators Whitfield Diffie and Martin Hellman were behind public key infrastructure (PKI), widely used to secure the Internet.
Play dates with data
Australia, too, has played its part. The widely used open-source 'rsync' algorithm, a fast file transfer and synchronisation tool, was co-invented by ANU graduate Andrew Tridgell. Danish brothers Lars and Jens Rasmussen, with Australians Noel Gordon and Stephen Ma, pioneered algorithms to map geo-locations; their start-up, Where 2 Technologies, was bought by Google where their innovations evolved into Google Maps.
Algorithms have become more advanced as computational power and complexity has evolved.
Advanced algorithms are transforming industries such as health care, education and space exploration. They look for everything from specific patterns in human genes in the quest to find cures for diabetes and cancer to cost-effective ways for rockets to deliver supplies to international space stations.
Good and evil: algorithms are used to both solve problems and sway opinion, says Ikram.
Both artificial intelligence (AI) and deep learning (DL) are advanced algorithms. AI algorithms enable computers (or any computing device) to mimic humans in solving problems. DL algorithms are subsets of AI algorithms; they use statistics to find patterns in data and allow us to imitate the way humans see, hear and even reason.
Are algorithms really our friends?
But algorithms are also blamed for many 'evils', from bumping a website down the Google rankings, to preferencing Facebook friends to cyber fraud and fake news. People, it turns out, are vulnerable to biases and manipulations delivered by social networking bots (sets of algorithms).
On social media platforms, people interact with others who share beliefs, preferences, likes. Algorithms facilitate those interactions. An algorithm has the power to sift through billions of recordings, billions of user commands, and assess the psychology of users. This data can be leveraged to sway opinion via social media or third parties.
False political news is disseminated more deeply and widely than false news about terrorism, natural disasters, science, urban legends and information about finance.
In 2018 it was revealed that millions of Facebook users' personal data had been harvested via a 'personality trait' app without consent by political consulting firm Cambridge Analytica and used in campaigns to win votes for Republicans in the 2016 US presidential election.
Digital days: interacting with Google, Facebook, a GPS and an ATM exposes us to powerful algorithm manipulations but Ikram says ad blockers can help us protect information about ourselves.
Special algorithms collected data from Facebook profiles (such as likes and posts) to estimate users' race, age, gender, political inclination and personality traits, and then targeted them accordingly.
Bad news travels fast
A study by Vosoughi, Roy and Aral, published in Science magazine in 2018, found that false political news is disseminated more deeply and widely than false news about terrorism, natural disasters, science, urban legends and information about finance.
- A new way of life: the great substitution effect of 2020
- Rising whale numbers a bonus for Sydney citizen scientists
The Cambridge Analytica scandal boosted the focus on cyber security. A team from CSIRO's Data61, University of NSW and UC Berkley in the US, which I was part of, analysed 283 VPN (Virtual Private Network) apps for Androids and found that 38 per cent contained security issues including malware, spyware, adware and data leaks.
There are, however, ways to protect the little privacy you have left: use an ad blocker so your activities are not tracked. Adblockers come in two flavors: browser extensions (also called add-ons or plugins) and standalone applications. Browser extensions can be found at the Mozilla Firefox, Chrome and Safari plug-in stores; and standalone applications are available from Google Play, Windows, and Apple iTune stores
And finally, only download apps that solicit the fewest permissions and least information about you.
Dr Muhammad Ikram is a lecturer in cybersecurity in the Department of Computing, Macquarie University. He is a member of the Information Security and Privacy (ISP) group at Macquarie, and a visiting scientist at Data61, CSIRO. He was assisted in this article by departmental colleague, Dr Alireza Jolfaei.
