Technology doesn’t always make our world a better place. Sure, thanks to the web, powerful search engines and devices like smart phones, a world of knowledge is at our fingertips any time of the day. But this has resulted in the near-death of the traditional pub quiz and I ask you if that is a price worth paying. OK, it might be but I’m here to tell you that we needn’t pay it! The pub quiz can live on and I’m going to show you how. For all their impressive capabilities, search engines today are still only dumb robots that basically work by trying to match a search query with the same textual phrases on the internet. They can’t (yet) very well understand the meaning of words and phrases, pictures, music or video. As such, there are a group of question types that they can’t handle. Here’s my guide to those question types with an example of each. Use them to create your own quizzes and have a go at answering my questions. Stick your answers in the comments section. Some are easy, some are hard. You can try Googling, of course but it won’t help you. 1. What Links? Contestants have to identify a common link between a series of names, places, words, numbers, whatever… Search engines can’t compute the myriad possible ways these things might be linked. Q1. What links: Daft Punk, Buster Rhymes, Kanye West and Alice in Chains? 2. The Odd One Out A cousin of What Links? and favourite round in Have I Got News For You, contestants have to identify a common theme in a series of names, places etc and find the one that doesn’t fit the pattern. Q2. Name the odd one out: Vinnie Jones, Julio Iglesias, Gordon Ramsey and Jason Statham. 3. The Next in a Sequence Another list that requires the contestant to first understand the relationship of the items before working out what the next in the sequence should be. Warning: this becomes Googleable once the relationship has been worked out. Q3. What comes next? Russia, Germany, Turkey, France, ? 4. What is More Likely? Questions on statistical probabilities are very hard for most search engines to understand, although Wolfram Alpha is gradually getting there. Other search engines will struggle to even realise they are being asked a mathematical question unless it is presented as an equation. It is technically possible to use search engines to find information to help you get to the answer but it wouldn’t be practical in most pub quiz situations. Q4. What is more likely if you live in the UK: being hit by lightning or dying in a plane crash? 5. Logic Puzzles There is no current way for a search engine to understand logic so these are very safe questions for quizzes. Q5. On a game show there are 3 doors with a prize behind 1 of them. The contestant must pick only 1 door which he does. The door remains closed but the show host opens 1 of the other doors to reveal it empty. He now offers the contestant the chance of staying with his original choice or switching to the other door. What should the contestant do to provide the best chance of winning the prize? 6. Media Identification Be careful here. Google Goggles and Shazam are 2 mobile apps that use image and audio recognition software to identify media. To combat this, distort your media in some way by adding noise or reproducing it in a different way from the original form. Q6. Name these movie posters:   7. Word Puzzles A merging of the logic puzzles and perhaps images, these can include the Dingbat-style word puzzles but also the likes of word searches, crossword clues and riddles. Q7. Name the phrase: HAVEHAVE + HOLDHOLD As I mentioned, if you think you know the answers, stick them in the comments below and I’ll provide answers next week. The rate of advancement in search engine technology and artificial intelligence means that these questions will probably be answerable by robots within the next few decades. I promise to write another blog post about beating the search engines then.

On advertising sales of $3.7 billion, Facebook made a profit of $1 billion last year. Its imminent stock market listing this year, where they are looking at raising $10 billion will take the companies market valuation to $100 billion which is 100 times its profits. That would make the firm, just seven years old with a 27 year old CEO and employing only around 2,000 people, worth about the same as Boeing, a 95-year-old aircraft giant with 160,000 workers. What makes Facebook worth so much money, when even Google who has a solid bottom line at $10 billion is trading at 20 times its earnings? Looking at external factors the US economy is recovering but at 2.5% forecast this year it’s slow and not enough to create such heady valuations. Nor does this valuation make any sense from the share market perspective which is trading at 12 times its earnings. Further investors are wary at the previous dot.com listings that haven’t really performed too well and are wary of paying massive premiums for Facebook shares. Have a look at the table below to see the performance of some of these companies which have been trading at their 52 week lows recently.

Let’s put aside economics growth figures and market multiples and look at Facebook in isolation as a social network platform. Here is a company with 845 million users, that provides an easy and cost effective way to for individuals to network and companies to promote their brand identify at a global level. This is achieved by humanising our online identities which were previously thought to be too dry and cold in the online space. As a result of our expressive outpour, Facebook knows where we are, what we like eating, when we are upset and the brands we consume. To compare this to some of the other main user engagement platforms, Google has a vast depository of user search behaviour, twitter knows users’ thoughts and opinions and YouTube has 60 hours of video footage being added each hour.

Facebook has almost all of the above and more information about us. Currently its advertising model is based on user demographic and the information users specify about themselves which can be targeted by advertisers. It could in the future get around user privacy issues and use an advanced form of dynamic user profiling to target adverts. Here is a scenario of how this could work: Tom is a regular Facebook user who shares pictures, comments and is part of different groups. During the run up to his birthday, Facebook could use his timeline data to determine who he interacts with the most on face book and things he likes. So if he has been part of the Spartacus TV show group, it could show his close friends bespoke adverts of Spartacus DVD from a retailer with a special discount if they bought it for Tom. That’s not just gold dust its diamond dust. Facebook critics, worry that with 85% of its revenues coming from advertising it’s not a very diversified business model making the upside relatively limited. In Facebook’s support, these same critics seem to forget that over 90% of Google revenues are from advertising and it’s doing alright! I feel that the case here for evaluating Facebook’s potential is not in its diversification of revenues but in its durability as a business and its ability to effectively monetise its 845 million users. I am quite sure that as long as people fly and Boeing makes safe and competitive products, it will be in business. However one cannot be so sure about Facebook’s future, when only recently we have seen another high profile social platform go under in Rupert Murdoch’s hands.