The full catastrophe of unmediated reporting has, by now, become depressingly obvious. At the heart of the problem is the ease with which the creators of junk information can reach a vast audience. The major delivery vehicles for this information, Google, Yahoo, Bing and Facebook deliver pages in such volume that the possibilities for human mediation are at most very limited. Instead, they rely heavily on algorithmic filters, including, but not limited to, Deep Learning systems.
One may have hoped that it was much easier, in the field of Science, to find and discard fake news articles. Sadly, today I discovered the naivety of such a view.
Google conquered the world of search by starting their company with a brilliant “page ranking” system, and then continuing by constantly improving the algorithms they use to present relevant pages to the searcher. One consequence of this approach is that identical search strings generate different results for every user, at each time, and at a given location.
Google also offers a news feed which is one of the most widely used news feed in the world. In July 2017 it ranked second according to ebizmba.com, garnering 150,000,000 unique monthly visitors if their numbers are to be belived. The News feed, presumably, also customizes its feed based on whatever information it can glean from the ip address of the person reading the news.
Today (March 4, 2018) the lead story in the Science section of my Google News was this:
The Daily Star article itself was also published today. It is unlikely that page ranking had anything to do with Google’s decision to show this story and put it first. It is more likely that Google remembered that I had recently been reading about Fermi’s Paradox (roughly: if there are aliens, where are they?) Whatever the reason, the story should not feature in the Science section. Unfortunately, unlike many other news services, there is no mechanism in Google News for any kind of human input (except, possibly, in a Google office somewhere).
Lets dig a little deeper into the article. Under the headline, appears the following:
A TOP scientist has accused NASA of covering up evidence of alien life on Mars – and he’s planning to prove it with a groundbreaking report.
The article is a perfect example of fake science news. The scientist they quote is not a “top” scientist by any objective measure. He has written a book or two. An Amazon review has this to say about one of his books:
Thumbing through this book at the bookstore, and seeing all of the NASA pictures, I got the impression that it is an authoritative treatise on life on Mars. Instead, once I read it, I found that it is a badly written conspiracy book. It seems that the author is convinced that the entire space science community is allied against Gilbert Levin, the book’s central character. The author seems to see coverups by NASA in every corner – all with the odd purpose of covering up what the author says is clear proof that the Viking robots discovered life on Mars. In addition to the conspiracy aspects of the book, it is horribly organized and jumps all over the place confusing this reader immensely. Someone should have hired an editor and a few real scientists to review this book before it was printed.
The headline of the article is, of course, classic click-bait. The “Mars mission” it refers to is not scheduled for another two years, hardly consistent with the sense of urgency implied by the headline.
The article is published in “The Daily Star” London, a British tabloid renowned, perhaps for many things, but not for the quality of their science reporting. Any thoughtful person with some knowledge about astronomy will of course dismiss the article for the nonsense it is, but the problem is that so many less informed people might read the headline, or, even worse, the article.
In the world before Internet news, the damage was limited to the people who read tabloids in Britain. But today, Google (in this case) has distributed this rubbish as a leading science story for the entire English speaking world.
Perhaps it is time to retire unmediated news feeds.