Hot off the press! Start your week off right, catch up on the latest.
Lack of imagination
Gartner has broken out its crystal ball to proclaim the 12 tech trends that’ll disrupt marketing! Bots, vocal search, IoT, AI, programmatic… OK, we suspected that. But what really grabbed my attention? More privacy-friendly data exchange (thanks to the GDPR)… Could this be a flicker of hope in the lack of imagination deplored by Nick Nguyen, VP at Mozilla? He evokes our lack of collective imagination as a true “tragedy of the web” in the face of search engines and social networks that “strive to give you a heavily filtered version of the web”. While Sir Tim Berners-Lee, inventor of the web, warns of the dangerous “myth that it’s too late to change the way platforms operate”, researcher Julien Falgas decries the way tech giants trap users in their navigational funnels and issues a wake-up call.
Researchers, communities, users, governments… it’s up to us to encourage initiatives that preserve an internet that’s thoughtful, welcoming of many voices, and that cannot be controlled by a small few. Take needle for example, a “contributive web navigation” project that’s open and distributed. Inspiring, isn’t it?
Ethics by design
And now for a perfect transition to this question: Can ethics and the search engine business model be compatible? In theory, it seems simple. Freedom of expression should be guaranteed for both search engine users and for those who index the results. OK, but what about applying local legislation? What about fake news, illegal content, or legal content that’s violent or obscene… What about questions of business partnerships and targeting… Who decides all this? Search engines intrinsically collect information on their users, even gathering information based on the questions they ask. Unlike other search engines, Qwant “doesn’t want to know what you’re doing” to make money from it. No data collection, no cookies, no history. Guillaume Champeau, Qwant’s ethics and legal affairs officer, talks about ethics as “a constructive constraint which enables us to stand out from the competition, to offer a different type of service, to hold different kinds of values in a world that needs diversity.” But the Californian tech giants don’t seem too concerned with these questions, between their funding from Saudi investments, abusive use of data, and software used by border agents to process and separate immigrant families at the American border. Who will teach Silicon Valley to be ethical?
What is a graph? We see and use graphs so often that we nearly forget what makes them appear before our eyes. A graph is data. The mathematical definition of a graph speaks of its nodes and edges. But essentially, a graph transcribes a set of relationships between different entities. Technology has enabled us to considerably accelerate our usage of graphs, but difficulty remains in finding solutions quickly (welcome to the world of NP-complete problems). In the real world, graph research can, for example, serve as a foundation for algorithms that optimise garbage truck itineraries (to reduce useless commutes and pollution). And since we love both history and dataviz, check out these 5 visualisations, each noteworthy for their time:
- The cholera map of London by John Snow (not that one):
- Hans Rosling’s visualisation displaying the relationship between income and life expectancy:
- The link between temperature and the French invasion of Russia, by Charles Minard:
- Florence Nightingale’s sunburst diagram illustrating causes of death (notably due to preventable or mitigable infectious diseases) during wartime:
- Correlations between former civilisations and their populations, from Joseph Priestley:
Last (but not least) click
Advertisers are still racking their brains over the infamous question of attribution… the eternal debate over who (or what) deserves all that glorious credit for conversions: the last click? The user’s prior interactions? How far back should one go?… Here’s what we do know: the longer the sales cycle is, the more difficult it is to measure the impact of channels from the early stages of the customer journey. Current attribution models tend to downplay the influence of organic channels in favour of paid channels. Keep in mind that your site-centric measurement (via your analytics data) remains crucial for understanding the relationship between all these channels.
It would appear…
…that it’s possible to earn $100 per hour to play with puppies, but take note that real skills are required… #MakeBullshitGreatAgain.