Alex Kelleher’s Blog

The concept that “information wants to be free“  has been around one way or another since the dawn of time.   News at the dawn of time (whenever that was) was most likely passed on by word of mouth – town criers, or neighbours in the cave next door.

These days it’s being touted as, amongst other things, the reason for the downfall of the newspaper industry.   As in the table below, only ONE of the UK national papers (the Daily Star, more of a celeb-paper than a newspaper) is up year-on-year in terms of circulation, and some are falling by as much as 15%. 

So what does “information wants to be free” actually mean?  As far as I can tell, the raw data that is the news (i.e. what’s happened) isn’t something the general public are any more willing to pay for.  As long as there are ways to subsidize the collection (or recycling) of news online, some or all of it will be offered free, and therefore no-one will be willing to pay for any of it. 

Some commentators bite back with comparisons like “Information doesn’t want to be free any more than gasoline wants to be free” – which is patently nonsense.   If I could find a business model to make gasoline free, gasoline would be free, almost overnight, everwhere. For instance, if I when filling my car up I watched 5 ads, filled out a long personal questionnaire, and did my weekly online shopping with an advertiser provider – there’s my free gasoline, job done.

The problem is in part that information covers a huge array of things: from “what time is it?” to “which stock should I buy” – and a whole complicated spectrum in-between.  And no-one can draw a line to say where the charging should begin.

The standard response of newspapers is that they don’t actually present the news – they present a mix of news, analysis, and “premium” news such as financial information or sports results, and so on.   It’s a form of entertainment.  Here’s my take on that:

• I reject the “premium” notion, as all news suffers from the same free-effect: if someone somewhere is willing to offer it for free, everyone everywhere expects to get it for free.  Even the FT online has only got around 117,000 paying subscribers around the world
• Analysis indicates that someone (or some people) who are smart have spent (chargeable) time processing the news, and coming up with useful views, slants and summaries - that we ourselves couldn’t or wouldn’t want to do.  It might be valid, but then 1billion blogs, including this one, do the analysis piece (if you can call it that in my case) and give it away.  Citizen journalists take the photos (more quickly).  It may not be as high quality – but then neither is the video on YouTube, and look where that got it
• The entertainmentpiece is probably the piece the Daily Star has got right – light on the news, heavy on entertaining, light and untroubling content

So what can the newspapers actually do?

1. Put up a “pay wall” so high that people actually turn back to cheaper print: As in the Newport Daily News (http://www.newsweek.com/id/214607).   I think this is a crazyand short-term idea.  In fact a pay wall of any kind is a crazy idea.  The common sentiment is that if Murdoch does it, everyone will have to do it.  But that’s bizarre thinking.  If Murdoch and everyone else puts up a pay wall, I’m launching a free newspaper, and boy am I going to get a lot of traffic.

2. Find smarter ways of monetising content – contextual ads are okay, but tighter links into other ways that consumers spend online is smarter.  This is my favorite approach.  Newspapers still (just) have a great mixture of brand, loyalty, customers and quality editorial staff  – and can churn out the content to support the revenue models to keep the industry alive.

3. Forget chasing the “next big charging thing”like ereaders.  Ereaders will very soon (if Apple has its way, by next year) be always-on web terminals.  Same story, same argument, and same free

4. Create entertainment products, like the Daily Star.  That seems to work, still.   But that may in part be because the audience of the Daily Star has yet to become as fully web-enabled, and in time too that advantage will erode

5. Realise that free means “free to the consumer”, not “completely without revenue”

In conclusion, I’m on the “information wants to be free” side, but absolutely don’t believe that means that there won’t be revenue models or a place for what we today know as newspapers.  In fact, as information power-houses, they’re really well positioned to take advantage of digital.  If only they’d just focus more on that, and less on lamenting the demise of paper and ink.

admin Web, news

Having just been to WPP’s Stream 09, and spoken at CTAM’s Eurosummit last month, I’ve been exposed a fair bit to a topic that affects both the cable (TV) and online industries fairly equally: video.  And, more importantly, how to monetize it.

YouTube generates billions of hours of viewing, and still are struggling to break even.  I’ve talked to people at YouTube, and clients of YouTube, and the frustration is the same: no clear way of applying advertising to the medium that both brings benefit to the brand, and to the user.

My 2cents worth on that a bit further down, but before that, something that just jumps out at me: Video is video is video!  There will be (at latest within 2-3 years) no distinction between online, digital TV, cable.  If (as someone else at Stream pointed out) I want to watch Family Guy, that means I want to watch Family guy.  What screen it’s on, what size that screen is, and what the platform is called doesn’t matter to me one bit.  What matters is watching the show – and soon (by which I mean very soon) I’ll have a set-top box under my TV that is “online”, no different from the box under my ‘lean-forward’ PC screen.  Forget the puny attempts by Panasonic et al to put some widgets onto a TV screen.  I mean full online access, with some optimisation to compensate for sitting 8ft away. 

What am I participating in then?  Online video?  Video on Demand?  Digital TV?

No, just video. So there is no “online or TV” debate, to my mind.  There’s “video”, that’s it.

Back to monetization: I did make the point to the YouTube guys that Google hasn’t yet done what it did so smartly with search.  There, it turned advertising into a user benefit (GoogleAds are of course frequently more relevant/useful than the organic results and generate up to perhaps 40% response from the audience).   And the format matches the context (short text snippet search results). 

With YouTube, Google seems to be forgetting its drive to innovate, and just wanting to copy TV advertising (pre-roll, post-roll and so on), or slap something up (annoying overlays).  The debate seems to be about how to annoy users least (by only pre-rolling, keeping the ads short and so no) rather than how to help users the most.  Helping users the most to my mind will help brand advertisers the most.

What’s really needed (especially in short form video, long form can probably carry TV-style advertising for a while) is a new, game-changing way of monetizing the content.  I spoke to quite a few people at Stream about this, and I’ll bet something will emerge soon.  As a thought-starter I regularly mention Net-a-porter (and have talked to Mark Sebba the Chief Executive recently about this and other things) as a great, simple example of smart monetization of content.  In their case, it’s like having a copy of Vogue magazine where women can click  on the clothes on the models, and buy them immediately. 

Perhaps there’s something in that approach for brand advertisers and video?

admin Television

So the whole world is going personalized.  Even in medicine, which I guess most people would think is personalized anyway, it’s the Next Big Thing.  Of course, it’s personalized to a degree in that your doctor will use their judgement in what you might have, but if you have the same disease as someone else, you’re likely to get the same drug and treatment.

That’s changing.  In fact, Pfizer’s DX Division Head says that personalized programs are “happening in the development and commercial stages right now”.  And that’s from the world’s largest pharmaceutical company.  There’s $110bn of company valuation they have to protect in the future, so you can bet they’re spending some good money in this area.

In the UK, the government last month released a note on personalised medicine, covering the current movement away from the “One size fits all” approach.  This note suggests that genetic differences between us can account for up to 95% of the variation in how we each respond to drugs. 

Really what we’re talking about here is a more granular approach.  Your medical, family and life history is combined with an ever greater number of lab test to build up a picture of you as different from another person with the same condition.  Strictly speaking, this isn’t “personal”.  It doesn’t mean that as an individual, you are unique, and will get a treatment no-one else will get. As the UK government report says, “Personalised medicine tailors treatment to patient subgroups“.

Of course it just isn’t possible to test new or existing drugs against every single person who could take them.  So the best that can be expected right now is to know that a drug works for “someone like you“.  One day, computer modelling of the effects of a drug on me as an individual (which would be my DNA plus my current health state, and changes that the environment and life have made – in fact a digital capture of the biology of me right now) will of course be possible – delivering truly personalised medicine. I’ll then pop a pill that would work for no-one else, and bingo – cured.  Until then, I plan to avoid all germs, and eat only aloe vera.

admin Technology, medicine

A couple of years ago I helped a friend (@JulianFriedmann who is a UK film/book agent) launch a site  for screenwriters, TwelvePoint.  I’ve talked to him many times over the years about storytelling.  In fact I contributedpart of a chapter to his book “How to Make Money Screenwriting“.   That chapter talked about the physiological/psychological effects of watching a movie, and what screenwriters might do with that information.  For instance, did you know that if you sympathise with a character on the screen who is running, your leg muscles will twitch in sync to their running?

Of course one of the basic truths about screenwriting (particularly US screenwriting) is the common structure of movie screenplay.  There’s a focus point in the narrative around page 15, a turning point around page 30, another focus point around midway, and a final turning point about 15 pages or so from the end (each page translates roughly to a minute).  Analysis shows that the vast majority of US movies are based on this formula – next time you watch a movie, check it out, it works.

This, and countless books and research, all point towards stories having a basic, common structure.  So it made perfect sense to me when I read about people looking at the concept of an ”Automyth“.  People (mainly in and around computer gaming) are looking at this, so there’s a lot of reference to fairy tales and ghosts and wizards and everything Terry Pratchett-esque.  For instance someone has created an online “Fairy Tale Generator” at http://www.brown.edu/Courses/FR0133/Fairytale_Generator/gen.html which does an okay job of fixing up a tale based on a bunch of selected inputs like “victory”, “rescue” or “difficult task”.

True, mainstream Automovies or Autobestseller novels are a little way away, it seems, but can it be that long before the first computer-generated bestseller hits a screen near you.  The technology to create the visual image, spoken word and automated music are there already.  All we’re missing is a piece of software that can generate a decent story.

UPDATE: Julian just sent me the following tweet: @alexkelleher I am the friend Alex helped; great blog in light of claims the 3 acts doesn’t work. Read Bettleheim Uses of Enchantment.

admin AI, software

Apparently. 

Craig Venter, one of the guys behind the Human Genome breakthrough, has just claimed that after a decade-long effort, his instituate is that close to creating artificial life.  Here’s how you can do it at home:

Step 1:Build the entire genetic code for a single bacteria
Step 2: Insert this code into a “host” cell
Step 3: “Reboot” host cell

Until 2007, Step 3 is where the process got stuck, the resulting bacterium just wasn’t viable.   Just recently, Craig and his Institute think they’ve cracked it using “methylation”, a way of protecting this new DNA from the cell’s defence systems. 

This breakthrough could give biology and medicine some great tools – in fact Venter is working on bacteria that transform coal into cleaner natural gas.  Pretty good for a guy who’s already helped us understand our own DNA, and is on a quest to “put everything Darwin missed into context” through his oceanographic surveys.

admin Genes, Technology

The New York Times chose to lead this story by pointing out that “the unemployed have more time for leisure and socializing“.   Yes, that seems pretty likely.

But in any case the interactive graph they created from their data is a pretty interesting breakdown of how different groups spend their days, by time of day. 

It’s full of “obvious” information, like men seem to watch more TV and do less housework, but it’s intriguing to play around with the different groups, they’ve certainly tried to analyze the data.  Here are some facts that I’m sure you wanted to know:

• Women shop for an average of half an hour a day
• At midnight, 2% of all Americans (which equates to 6m people) are working
• At noon, 4% of all Americans (12m) are asleep…
• 25% of people spend more than an hour travelling to work
• At 8:50pm, 2 fifths of people are in front of the television

More evidence that we’re a pretty predictable bunch of beings…

admin data mining, visualization

“Robots that can decide where to kill, who to kill and when to kill are high on all the military agendas,” according to some recent commentsby AI professor Noel Sharkey. 

I’ve talked previously about the early steps machines are making towards autonomy, and combine this with weaponry and you’ve got yourself the plot of a couple of dozen Hollywood movies.  Isaac Asimov’s 3 rules for robots may be needed sooner than we thought! (In case you hadn’t seen them before, they are:

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey any orders given to it by human beings, except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.)

Now that’s all very well for I, Robot, but autonomy in robots will inevitably be reached well before they’re intelligent enough to use that autonomy, or are capable of sticking to rules.  Relatively simple AI will be used in machines that potentially have access to damaging/lethal technology.

Think that’s impossible?  The headline above is real, and to quote the article “the robot suddenly came to life and grabbed a tight hold of the victim’s head. The man succeeded in defending himself but not before suffering serious injuries.”

Prof. Sharkey is calling for debate on autonomous military robots, and it’s timely – but probably won’t happen until one of these military robots does something “wrong”, such as a friendly fire incident.  Be that as it may, A.I. (or at least highly trained autonomous machines) is of course being developed on in a lot of diverse areas.  And given military budgets, it’s quite possible some of the first properly autonomous and intelligent machines will carry a weapon…

admin AI, Military

Okay, there are some other benefits of bacteria (as the TV ads for live yogurts keep reminding us), but I’m talking about something altogether more powerful: bacterial computers.

A group of scientists in Missouri and North Carolina have used bacteria to find solutions to an (essentially mathematical) problem: navigating a group of dots so that each dot is only crossed through once, and then returning the first dot – as in the image on the right.  It’s called the Hamiltonian Path Problem, if you’re interested.

So, it’s not quite Windows 8 yet.   In fact, the importance of the study has been questioned variously, but the reason I love it is that it’s using living organisms to find correct answers to mathematical problems, by simply chucking a designed molecule at them.  We’re starting simple (and slow), but then so did computers.   Anyone remember the ZX81?

Some other developments in this space over recent years include:

• DNA computers that can diagnose cancer
• Leech neurons that can do sums
• Molecular devices that only treat cells that need it

So why is this all important? Solving hugely difficult computational problems like predicting the weather or curing complex diseases has two approaches now.  The first is “synthetic”: computers, silicon, small machines with cogs and motors, and the like.

The second is “organic”: recognising millions of years of evolution, and saying – how can we use that?  Given that a lump of biological material can create something as fantastically capable as a human brain, why not harness the same systems to do other things?

And that’s where the genius is: short-cutting innovation by harnessing what’s already out there.

admin AI, Technology

My first encounter with Mandelbrot was when my Dad coded it up on a Commodore 64 (yes, let the nostalgia begin).  The math didn’t mean much to me then, but the pictures were nice.  Or at least I thought they were, they looked something like this:

The thing about this type of maths is that it’s “self similar” – when you zoom in, you get the same shape over and over again, as in the animation below (courtesy of Wikipedia).  That happens in nature in everything from flowers to skin wrinkling.

So I was pretty intrigued when I found this site, which uses similar math to generate images – and suitably updated for today’s digital eye:

That’s it.  Mathematical eye-candy.

admin Math

Every so often, a story comes a long within the field of psychology which underlines just how far we’ve still got to go. 

In a study at UCLA, subjects were asked to do one of 8 different tasks – from saying whether words rhymed, to counting the number of tones they heard.  So – very different, varied tasks. 

Investigators using fMRI (functional magnetic resonance imaging) – the very cutting edge of brain scanning – and then had to guess from the results of the scans which tasks the subjects were doing.

Chance alone would get 1/8 or 12.5% of guesses correct.  They managed 80%.  Pretty good – but let’s just remember that they were looking at tasks that were very (very) distinct.  And they still got it wrong 20% of the time…

How did they do it?  With machine learning math (the same kind of math that we use at Cognitive Match).   And yes, it’s a bit of a good result for this field of technology.  But far from perfect, and very far from mind reading.

Of course, all this means is that fMRI and the way we interpret the results from that technology could be better.  Much better. 

But more importantly – maybe it’s just the wrong thing to do (which is my take on it).  It’s a bit like working out whether the stock market is up, based on satellite weather images.  The brain just doesn’t work based on large areas of electrical activity on a scan.  It’s complex, enmeshed, multi-layered, abstracted and part of a body of integrated biological mechanisms.   And I hope that’s clear.

admin Profiling, Psychology