Wikipedia-backed Wikitribune won’t fix fake news

wikipedia wikitribune truth

Those past months, we’ve been hearing about Jimmy Wales’ ambition to create Wikitribune, a system that analyzes potential fake news to sort out the truth from the junk. Wikitribune claims to deliver exclusively “evidence-based” information, or put more simply “true news”. While I have a lot of respect for what’s been built over at Wikimedia & co, there’s one thing that Wikipedia is not: A reliable source of true information when it comes to controversial topics.

On Wikipedia, in case of a disagreement among the contributors of an article, two top principles are supposed to self-handedly manage the feud :
NPOV (neutrality of point of view): NPOV is reached by bringing up all verifiable POVs on a given topic, instead of working towards a more elaborated way to filter the controversial info.
Consensus by majority: In case of disagreement between the contributors of an article, a consensus must be reached in the discussion page before going any further. Thus the biggest group represented wins, regardless of the veracity of each party’s arguments, because it forms its own consensus (like a leading majority in politics).

Therefore, on Wikipedia, the biggest group of opinion represented can win over any battle of opinion because it will have the lead on controversial editorial decisions. And since NPOV is defined as nothing more than a patchwork of POVs, the winning majority will have the entire freedom to define how that’s going to be applied. In some cases, the minority abandons the field and the Wikipedia article becomes nothing more than a feast of biased and misleading information.

Needless to mention that when a leading party on Wikipedia also has connections in the press, the barriers of Wikipedia against distorted or fake news become useless (because if the press says it, it’s true). Wikipedia is a great vehicle of knowledge, but it certainly is not a conveyor of truth in any way. Therefore, there is no logical reason for Jimmy Wales to launch Wikitribune, because he has nothing to back it up with.

Source of featured image: The ‘Undue Weight’ of Truth on Wikipedia, which is also a great article to read on that topic.

calvin hobbs history

Content is Go(l)d

content is king bill gates

Google Search, Amazon, Wikipedia, what do these companies have in common ? They’re web giants, they were created at least 15 years ago, and they’ve all showed a very conservative approach to web design.

While social-related services such as Facebook, Youtube, Twitter have greatly evolved in terms of design in the past ten years, Google, Amazon and Wikipedia have showed little aesthetic improvement. Obviously though, their services have greatly involved:

  • Google Search’s algorithm became so sophisticated that the smartest engineers are not able to fully understand it, and Google is now a global web player that dominates the online industry way beyond search;
  • Amazon enriched its online sales power through new features and developped a winning online retailing strategy;
  • Wikipedia developped a great, vibrant community, it now exists in almost all the languages used by humankind, and it contains tens of millions of well-documented pages, all-the-while maintaining its free and open policy.

But that stuff lies under the hood for the most part. What the user see hasn’t changed much in the past 15 years. Which brings the question : Is a plain, generic design the key to a stronger brand on the web ? If an entrepreneur had big ambitions with a new kind of online service, should he try to make his design as groundbreaking as possible, or should he stick to the Craigslist-like list of links and plain text ?

A Google engineer would remind me at this point that Google spends a whole bunch of money testing its design on user groups to ensure it gets it right. And while you can hardly argue with scientifically-proven facts, it seems that web design has made lenghty progresses in the past 15 years (HTML5 held the promise of a greater, more interactive web for example), thus it would seem logical that those big companies should have followed this lead to remain attractive. They haven’t, and yet they keep prospering.

Content is king; Design doesn’t matter, content does; Build great content and they will come. You better believe those statements are true as Google, Amazon and Wikipedia are proving it right. These companies are showing that users are attracted and retained through simple interfaces and easy to understand ergonomies. So if you think you have a great idea for a new web service, try to design it as plain as Google, Amazon or Wikipedia. If it looks useless and leaves you clueless, then there’s a great chance that your idea is not that great after all.

Powerset’s Launch and Sell Strategy


Powerset today announced on their blog that the transaction with Microsoft has been finalized. The cool natural language search engine isn’t a wriggling fresh startup anymore, but has made the Darth Vader’s move towards the dark side of Microsoft’s heavy search infrastructures.

So Powerset isn’t one of the hype search startups of the Valley anymore. There has always been much ado about Barney Pell’s ability to buzz and sell startups, but I think that Powerset’s story is a case study that should be taught in tech economy classes.

First, consider the unusual amount of press they got in Techcrunch since August of last year:

This is more than 15 articles featuring Powerset in a year. Web startups usually get an average of 0.4 reviews a year on Techcrunch. In the case of Powerset, their news got covered when they were looking for a new CEO, when they made a case study with Miss South Carolina, and 4 articles were about their acquisition. Rather unusual… Undoubtedly, the PR firm contracted by Powerset did a good job creating compelling news about the company. Nonetheless, not all their clients get such good coverage. Powerset’s strength lied in their intrinsic buzz strategy:

  1. The company is a search engine – it positions Powerset as a virtual threat for Google (i.e. as an opportunity for Microsoft).
  2. Their first product is a Wikipedia search tool – it positions Powerset as a portal for one of the most popular Websites of all times.
  3. The search engine focuses on natural language search queries, a fantasy world – As Lorenzo Thione states in the interview above: “If you can crack that nut of understanding human language, with algorithms, with computers, then you open up the door to something that has been part of the collective imaginary for a long time”.
  4. The Founder’s got a great deal of startup experiences – Barney Pell has an impressive resume in the fields of search and VC activities.
  5. Powerset is part of the Silicon Valley’s cool kids gang – they co-organize exciting events for the local tech community.
  6. They get social media coverage – I don’t know what is their deal with top bloggers, but it works.
  7. All the factors above creates a bubble around the company’s name that consequently over-values the initial search engine project.

HyveUp – Tristan Harris – Apture


Apture is a new tool for bloggers that allows “content creators the power to find and incorporate relevant multimedia items directly into their pages” by adding links and small navigator windows to pages and posts automatically (CC Weblog).

This description of Apture couldn’t be more exact. I am a video producer, and until today, I had not found a way to let readers access and watch all of my videos without leaving that same page you are on right now (converting readers into viewers is tricky). With Apture, setting up one of those links is a breeze. Simply select directly on your homepage the word(s) you want to link to, and Apture automatically fetches for you related Youtube videos, Flickr photos, Wikipedia articles and WashPost content. Out of this automated pre-selection, you the publisher select what it is that you really want to link to.

Apture opens up the door to a new form of navigation system, where content is universal thanks to CC-license and Wikipedia’s GFDL license. As of today, small site publisher can try Apture for free, and the startup offers a revenue-sharing model for larger media sites. As far as I am concerned, Apture is one of the most innovative publishing solutions I have heard of and adopted so far. A lot of startups think that 3D is the way to tackle the flat Web symptom. Apture actually is a form of 3D, as it adds layers of content to your site. However, 3D is a little sci-fi and heavy on your bandwidth. Apture is down-to-earth and nice with your connection speed. Is this the Web 3.0 everyone is fussing about?

There is much more to be said about Apture, so read those great articles below:
With Apture, Hyperlinks Get Rich Media – Giga Om
Apture: Enrich Your Site’s Content – Brown Thoughts
Apture drags online news experience into the 21st century – Venture Beat
Fantastically cool code to watch – Lawrence Lessig’s blog

HyveUp – Lorenzo Thione – Powerset


Human languages are usually referred to as natural languages, and the science of studying them is linguistics. A common progression for natural languages is that they are first spoken, then written, and then an understanding and explanation of their grammar (according to speech) is attempted.”(Wikipedia)

Last week, I went over to Powerset’s offices in San Francisco to interview Lorenzo Thione about Powerset. Powerset is a search engine that analyzes search queries and retrieve results based on natural language algorithms (and not only keywords). The search engine is still in private beta, and is expected to open to the public this year.

The buzz says that Powerset is seriously challenging Google’s search technology. For now, the private beta happens on Powerlabs,where a community of users participate to the foundation of what could become the next consumer search engine leader.

Powerset will release the Wikipedia search tool in the next few weeks. Lorenzo explains that Wikipedia is a great environment for Powerset to fine-tune their technology on one of the Web’s most exhaustive and diversified source of information. Expect a lot of press coverage on Wikipedia search.