Can strong consumer data protection be a disruptive innovation strategy? Why is pi so interesting? Want to teach yourself computer science?
An academic paper has heavily criticised the data sharing deal between Google's DeepMind and the Royal Free NHS Trust for it's lack of transparency and patient involvement. The controversial deal was signed in 2015 to develop the acute kidney injury alert app Streams.
-- In subsequent sharing agreements between DeepMind and other London medical institutions, the data was anonymised before being transferred and the purpose for use clearly defined. DeepMind have argued that the data is being used for "direct care", as opposed to research - making it identical to many contracts between NHS trusts and clinical IT system suppliers. However, the transferred data includes the health records of 1.6 million patients - most of whom have never had an acute kidney injury.
Journalist Adi Gaskell, who has written numerous times about the benefits of pooling health data, emphasises that "it’s crucial that the process is done correctly and transparently." Gaskell also points out that there is the risk that DeepMind might gain a de facto monopoly on patient data - "which would be terrible news for providers, patients and the wider ecosystem." 🏥
Tim Berners-Lee published an article explaining the three things we need to change to save the web. Notably, Berners-Lee stated that we've lost control of our personal data. This was evident this week through the leak of a database acquired by Dun & Bradstreet(D&B) which contains more than 33 million records - including US government departments to large corporate clients. 😳 Importantly, the database includes personally identifiable information - such as full name, email and job title.
Troy Hunt, who runs breach notification site Have I Been Pwned, obtained the very carefully curated database and analysed the records to further iterate Berners-Lee's point around control of personal data. Interesting read. Hunt argues that it looks like the kind of database that would be sold to clients - further highlighting the numerous privacy and security issues around selling and moving data. ⛔
Can strong consumer data protection be a disruptive innovation strategy? Francisco Costa-Cabral argues that rather than "using data to compete on a product", companies can differentiate themselves and get ahead in the market by "competing on data protection" - such as the success of WhatsApp offering end-to-end encryption. 🙏
The French government have opened up their unemployment services database to non-profit big data company Bayes Impact. They want to build "a free, open-source software platform" to match people's skills to available jobs. Interesting use case for social good. 🤔
Describes a perfect circle - used in any formula that includes some form of repetition.
Appears random - equal chance of any digit between 0 and 9 appearing.
Looks random - mathematicians have computed pi to 10 trillion digits and seen no evident pattern.
But isn't - pi is mathematically fixed - any word that you can think of, when encoded in numbers, would show up in pi - even pi itself.
Since 2013, Martin Krzywinski has been producing visualisations to celebrate Pi Day: "the art is meant to awaken emotions about math and start conversations about numbers and randomness". Check out his 2015 treemap illustration of pi below:
Pi Day Graphic by Martin Krzywinski, via Washington Post.
Also, Martin Krzywinski re-imagined the numerical digits of pi as physical masses in 2016. Love this graphic. + Check out the rest of Krzywinski's visualisations, including for Pi Day 2017, here.
Slavomir Kaslev presents why types matter by showing what you can do with them. 👍 + QOTD: "Bad programmers worry about the code. Good programmers worry about data structures and their relationships." - Linus Torvalds.