Fortnightly Mailing

How algorithms manipulate the market

In September 2015 I took part in a symposium about "digital capitalism" organised by CRASSH (the Centre for Research in the Arts, Social Sciences and Humanities). This culminated in an evening talk by Dan Schiller - Digital Capitalism: Stagnation and Contention? - with a main argument that:

in contrast to the unabashed triumphalism that greeted the rise of the Internet as a pole of growth during the 1990s, today we are living amid both persistent economic stagnation and escalating political contention over the structure and control of the world’s information infrastructure.

Last Friday I went to a related one-day conference - The Power Switch: How Power is Changing in a Networked World - also run by CRASSH.

One speaker – competition lawyer Ariel Ezrachi – gave a particularly eye-opening talk. (You can get a flavour of what some of the other speakers said in this 35 minute Talking Politics podcast, led by David Runciman. At some point videos of some or all of the event will (?) be available.)

Here’s the gist of Ezrachi's talk.

1. A popular premise is that the Internet is a blessing when it comes to competition. Except that "The invisible hand of competition is being displaced by a digitalized hand".
2. Monitoring of the market by the platform providers means you see what they want you to see, as in The Truman Show.
3. We think we are savvy because we ignore targeted adverts.
4. But it is much harder (impossible?) to escape the effects of dynamic and personalised pricing.
5. Systems know your location, how long you’ve been pondering a purchase, the path you took. You get offered lower prices if you first searched, than if you came direct. If you did not conclude a transaction you may find the prices falls.
6. You may be offered higher prices if you searched using a Macbook than a cheap Android tablet; or from a wealthy location. If you are a first time buyer you may get a lower-than-normal price: even lower if you seem to be a rich first time buyer (because getting a rich person’s sign-up is worth more than a poor person’s).
7. The net effect of this is a hidden and unrecognised anti-competitive interference in the operation of the market, in which the house always wins, and where there is a real problem of asymmetric information between buyer and seller. (Ranking of search results to put cheaper options lower down is another example of asymmetry. As is the way Uber operates; as will be the way that your “digital assistant” operates, using data about you to shape the prices it finds for you.)
8. In a shop you pay the list price or less; if you buy on the Internet the price will be the base price or more.
9. Users of mobile phones are stuck with only two platforms: Android and Apple, sitting underneath all mobile apps; and although Facebook is itself a super-platform, it is beholden to Google and Apple because the latter control the platforms through which mobile users access Facebook.
10. So, there is now an enormous concentration of power over the market – a new kind of power – in very few hands, with big scope to undermine new entrants, and reduced opportunities for disruptive innovation. (Thus Peter Thiel’s "competition is for losers".)

Summary. A fundamental change is taking place in the nature of market competition, with markets invisibly and continuously manipulated by bots and algorithms, making competitive pricing an illusion; and with power over the market infrastructure shifting (already shifted?) into very few hands, there are big and poorly understood risks to economic and overall well-being.

FE Area Based Reviews should start by making an assessment of need

[Also on LinkedIn. Header image added 3/10/2015. Small edits made 4/10/2015. Addendum from evidence by Martin Donnelly and Peter Lauener to the 19/10/2015 Public Accounts Committee added 20/10/2015]

Until 2002 I was employed in The Sheffield College. For the last seven years I have been a governor there. The college is a big urban FE college spread across four main sites, with a turnover of over £50m.

In 2000 I was a bit involved in The Sheffield Review, after The Sheffield College was put into Special Measures by the FEFC and the then Education Secretary, David Blunkett.  Two Governors were "imposed" (Bob Fryer and FEFC's Dr Terry Melia) and George (now Sir George) Sweeney was parachuted in as Principal.

Here is a link to the executive summary of The Sheffield Review: http://dera.ioe.ac.uk/9912/. (I have a hard copy of the full review, but there seem not to be any publicly available digital versions.)

One very striking aspect of the Sheffield Review's method, which seems to differ from what is currently envisaged for ABRs, was that the FEFC's Terry Melia was very hot indeed on working out what the need in Sheffield was, and only then moving on to what provision was required to meet the need.

Of course the funding situation then was not as it is now, and is expected to be even worse after November's Comprehensive Spending Review. So funding will be insufficient to meet needs. And working out need in a LEP area such as Greater Manchester, or Sheffield City Region is a bigger job than "only" in one big local authority area. Nevertheless, ABRs ought, for moral as well as practical reasons, have that baseline assessment of numbers/need at an early stage.

From what I can make out, having attended the introductory meeting for college governors about Area Based Reviews, and having kept my eyes and ears open, it seems that ABRs will look at what there is on the ground by way of supply, and then move on to considering how that supply might be better and/or more cheaply provided. [Note that this approach is not the one that Government Officials described to the Public Accounts Committee on 19/10/2015. See Addendum below.]

The gap between supply and need would thus never be analysed.

This flaw in the ABR process (it is not the only one...) should be fixed.

Addendum - 20/10/2015

Extract from the oral evidence taken on 19/10/2015 by the Public Accounts Committee Inquiry into the financial sustainability in the further education sector. Emphasis added. For the full transcript go to http://goo.gl/w2It35.

Q50 - Chair [Meg Hillier, MP]: It sounds like it could be a bit haphazard. In terms of the future shape of the sector, FE colleges particularly have a capital asset and a physical presence, which constrains who they deliver to, to a degree, but also is an important local provision for people who may not be able to travel in the same way that people might do to university. Does the Department have no strategic oversight of what the general geographical spread should be of these institutions, and do you not have any alarms or worries about how area-based reviews may work or throw up mergers that might not deliver for all residents in a particular area?

Martin Donnelly [Permanent Secretary at BIS]: We are very concerned to ensure that there are available learning opportunities for people throughout England—in this case. Perhaps I could ask Peter to comment in a bit more detail about how we take that into account as we go through the process of supporting colleges and the area review.

Peter Lauener [Chief Executive of the Skills Funding Agency and the Education Funding Agency]: First of all, a bit of context. As Martin said, there has been a long-term process of rationalisation and merger in the further education sector since incorporation in 1993. When a merger happens it does not mean that buildings are necessarily closed, although there is sometimes a separate process of reducing the number of buildings in a particular area, if there are too many; but very often you get distributed leadership and management over a wider area, which produces savings and efficiency, and improvements in effectiveness.

As we go to the area reviews, the big challenge with those is precisely, I think, what you said—to start with what is needed for learners, for communities, for business, and then work back from that to structure. So it is not a sort of “move the deckchairs around”. It is what is needed in this area to provide the best possible service to the three groups—learners, the community and employers—and improve progression through to higher skill levels. It is a quite a challenging agenda.

Citizen Maths - powerful ideas in action

[Cross posted, with some very minor changes, from LinkedIn.]

Since late 2012 I have been closely involved (as "project director") in the creation of Citizen Maths, a free open online maths course at Level 2. I last wrote about it here just over a year ago. 

Citizen Maths is a free open online maths resource for:

• self-motivated individuals who want to develop their grasp of maths at Level 2;
• employers/unions who want to provide staff (or, in the case of trade unions, their members) with a practical and flexible learning and development opportunity in maths;
• colleges and other learning providers who want to give enrolled learners an additional or alternative route to improving their maths.

To make use of Citizen Maths, learners need access to (and knowledge of how to use) a desktop or laptop computer with a broadband internet connection.

Here's a four-minute screen-cast about Citizen Maths from a learner's point of view:

 

Who is behind Citizen Maths?

Citizen Maths is funded by the Ufi Charitable Trust. It is developed by Calderdale College, with the UCL Institute of Education, OCR, and with advice from the Google Course Builder team.

What does Citizen Maths consist of?

We’ve designed Citizen Maths to involve between five and 10 hours of study for each powerful idea. It it built up from:

1. short “to camera” videos and explanatory screencasts, by experienced maths tutors Paula Philpott and Noel-Ann Bradshaw; 
2. activities, tasks and other practical challenges, using

• applets that provide an onscreen manifestation of a powerful idea
• the Scratch programming environment
• standard tools like pencil and paper, and spreadsheets.

There are also frequent “low stakes” quizzes to help users check their understanding.

Why “powerful ideas in action”?

Citizen Maths engages people in familiar activity to reveal the ‘maths inside’, focusing on the way that maths has an immediate relevance to the problems we all of us have to solve every day. These problems could range from comparing deals and prices on groceries and creating a household budget, to understanding a payslip, creating sales forecasts, keeping track of savings and pensions, controlling a production process, or making political judgements. By putting problems in meaningful contexts, learners who do Citizen Maths will begin to grasp the power of mathematical ideas in action.

Which powerful ideas does Citizen Maths cover?

There will be five. During autumn 2014 we ran a proof of concept trial of Citizen Maths based on the powerful idea proportion. From mid October 2015 Citizen Maths will embrace, in addition, representation and uncertainty. From spring 2016 here will be two further powerful ideas: pattern and measurement. Here's a summary of the scope and importance of each.

1. Proportion is about mixing, sharing, comparing, scaling and trading off. It sits behind many aspects of everyday maths, for example when you are sharing out costs, or altering a mixture, comparing amounts, or scaling something up or down.
2. Uncertainty includes making decisions, playing, and simulating. It offers a way of thinking about uncertainty in personal and work-related situations, for example when making sense of risks to health, deciding whether to take out an extended warranty, or playing card games.
3. Representation is about interpreting data and charts, comparing groups. It recognises how much we are influenced by data and the presentation of data, for example in media reports of opinion polls, interpreting stories about health risks, or comparing our own household income to that in the rest of the country.
4. Pattern is about appreciating structure as in tiling, or knowing how to construct such structure. Pattern focuses on how mathematics can find and describe the regularities in both the natural and the man-made world, for example in the symmetries of animals and plants or in the design of buildings.
5. Measurement includes reading a scale, converting, estimating, and quantifying. It picks up on the importance of measures and measurement in everyday and working life, for example when dispensing medication, converting currencies or estimating the size of a crowd.

To find out more, go to https://citizenmaths.com/. There is also this Slideshare presentation:

On the long-term future of artificial intelligence

Minor edits made 19/8/2015

Stuart Russell is co-author (with Peter Norvig) of the very highly regarded text book "Artificial Intelligence - A Modern Approach".

This 30 minute talk provides a striking, accessible and ethically focused explanation of what AI is, where it is headed, and why its practitioners need to find ways of making AI "provably beneficial", if it is not to have, long term unforeseen and harmful consequences. (Cory Doctorow has more on the talk here.)

A ten year old interview

Reda Sadki reminded me about an interview I did by email for Epic (now Leo) in July 2005 in which I'd banged on about wanting to ban the term blended learning.

The interview was long gone from the Epic/Leo site. But the Internet Archive's trusty Wayback Machine had it, and all but two of the links still worked, at least after a fashion.

I re-read it, initially with trepidation, then with quite a bit of relief. Here it is. (I've fixed the dud links and added one to a review I subsequently did of The user illusion, cutting consciousness down to size by Tor Nørretranders.)

Q What's your INTEREST in learning/online learning?

I spent 25 years working in Further Education, teaching and developing TUC courses for trade union representatives. Through the TUC I got involved in pre-internet online distance learning courses, using a Swedish conferencing system called PortaCOM. I applied what I’d learned in the creation of LeTTOL, a web-based online course for teachers wanting to learn how to teach on-line – http://www.lettol.ac.uk/, which, several thousand learners later, won a National Training Award in 2003. My interests now center, through ALT, on establishing learning technology as a discipline, and learning technologist as a profession, and in the other half of the week mainly on helping organisations implement sustainable e-learning.

Q What interactive technology do you use and have at HOME?

Several radios and a telly. All four people in my household have networked computers, one of which is a Mac, and one of which is used for making music. My sons use iPODs. No Digital TV. No games machines. No self-filling fridge. I have and use a lot of books, which you could class as an interactive technology.

Q What stands out as your MOST EFFECTIVE learning experience?

A week training to be a trade union studies tutor. Extremely challenging. Plenty of feedback. Combining learning about a curriculum with learning how to tutor it. Reading “Inside the Black Box – Raising Standards through Classroom Assessment” by Paul Black and Dylan Wiliam. An in-a-nutshell summary of why giving learners timely and motivating formative feedback is the most important determinant of how fast and well they learn. http://www.kcl.ac.uk/depsta/education/publications/blackbox.html [Now available here: http://www.webcitation.org/6VELZxcop - SS 28/6/2015.]

Q What stands out as your LEAST EFFECTIVE learning experience?

A year training to be a further education teacher. Diffuse. Lacking in practicality. Thin on (useful) theory.

Q Any really NEW AND INNOVATIVE IDEAS out there?

When I see the word “innovative” my heart sinks, even more so when I see the words “really new and innovative”. This is because I believe in honing and improving ideas and methods which work, rather than moving to the next fad, and in e-learning there are a lot of fads. Of course the danger with this approach is that you can be blind to necessary or beneficial innovations. So, if pushed I would say that applications like http://www.jot.com/ which enable users to build Wikis without any special syntax are worth keeping an eye on, as are tools like http://search.yahoo.com/cc which finds content across the Web that has a Creative Commons license.

Q What do you want that DOESN'T YET EXIST in learning/online learning?

Machine translation! But this interesting piece about “The Google Translator” - http://blog.outer-court.com/archive/2005-05-22-n83.html - perhaps shows that something sitting in the background which enables people to converse with each other online when using different languages is not that far off.

Q Any views on the phrase and concept 'BLENDED LEARNING'?

The term provided a bolt hole for traditionalists wanting to defend face-to-face teaching against the encroachment of online learning.

Q Any views on GAMES in learning/online learning?

I trust my sons’ judgement that the value of games in learning is exaggerated. But I think I am probably missing something.

Q Any views on INTERACTIVE TV in learning/online learning?

In a previous role I helped develop “Keep IT In The Family”. This was a simple quiz – a game, even – to test a user’s IT knowledge, at three levels of difficulty, and to recommend suitable IT courses depending on the user’s knowledge. It was served from The Sheffield College and was freely available over the Internet, or to Telewest DiTV subscribers. At one point, judged by the number of users, Keep IT In The Family was one of Telewest’s most popular interactive services. That said, I feel that learning is a category of activity which normally requires learners to be able to concentrate, free from interruption, with a means of making complex inputs (currently using a keyboard). TVs typically neither have the necessary input devices, nor is a living room a conducive environment for learning.

Q Any views on MOBILE DEVICES in learning/online learning?

I’ve not yet read “JISC Landscape Study on the use of Mobile and Wireless Technologies for Learning and Teaching in the Post-16 Sector”. Certainly the pressure is now on content developers to make sure that content will run adequately on a wider range of access devices than just a PC or a Mac. And users of mobile devices are paying for data by volume rather than at a flat rate. So they may not thank you for media-rich content, even if it is educationally effective.

Q Any views on OPEN SOURCE in learning/online learning?

Open Source. I use Firefox and Thunderbird as my main browser and email client. Moodle, for example, is certainly presenting an interesting challenge to LMS vendors. But in 5 years time I think there will continue to be a “mixed economy” of software products in the provision of e-learning.

Open Content. Initiatives like MIT’s Open CourseWare - http://ocw.mit.edu/ - and the stunning W3 Schools web site - http://www.w3schools.com/ - show the power and significance of freely available e-learning content.

Q What's your favourite PHRASE/QUOTE/EPIGRAM in learning/online learning?

Because Jacob Bronowski’s “The Ascent of Man” was so influential, and because so many of his quotes make you think, I was disappointed to find that I’d been wrongly attributing “A word is worth a thousand pictures” to him, including the accent. It is still my favourite phrase in learning/online learning, mind.

Q Could you recommend a PIECE OF RESEARCH in learning/online learning?

Learning styles and pedagogy in post-16 learning: a systematic and critical review. This report, by Frank Coffield, David Moseley, Elaine Hall, and Kathryn Ecclestone, is freely available for download from the Learning and Skills Development Agency. It critically reviews the literature on learning styles, and it calls into question the way in which learning styles inventories are in widespread use, often with next to no evidence as to their validity. http://www.lsda.org.uk/pubs/dbaseout/download.asp?code=1543 [Now available here: http://www.webcitation.org/66qgBO959 - SS 28/6/2015.]

Q Could you recommend a BOOK in learning/online learning?

The user illusion, cutting consciousness down to size by Tor Nørretranders (ISBN: 0140230122). [Review here http://fm.schmoller.net/2007/03/16_bits_per_sec.html - SS 28/5/2015.] More about the nature of consciousness than about learning, but provides convincing evidence that the conscious mind is only able to deal with a tiny proportion of the data it receives - perhaps as little as 30 bits per second. The mind then creates a “media-rich” consciousness from this thin data-stream. We’ve evolved to interpret the sensually complex real world in an effective way; but that does not mean that our brains are good at effectively interpreting media-rich learning materials, which should hence be used (if used) with great care.

Q Could you recommend a WEBSITE in learning/online learning?

W3 Schools - http://www.w3schools.com/.

Q If you were to pick one CONFERENCE to attend in learning/online learning, what would it be?

ALT-C. Why? I work for the organisation which runs it. ALT-C has enough depth and breadth for an astute delegate to be able to plot a varied, interesting, and rewarding course through it. The booking deadline is 12/8/2005.

Q Any words/phrases/ideas you'd like to BAN from learning/online learning?

Word. Blended.
Phrase. Compelling content.
Idea. Digital natives and immigrants (which is not to say that Mark Prensky’s Digital Game-based Learning (ISBN: 0071363440) has nothing useful to say – both it and he have!).

Q Anything in learning/online learning that you strongly believed in, on which you have now CHANGED YOUR MIND?

I used strongly to believe that learning without some face-to-face contact between learners is unavoidably and badly second best. Thus online distance courses just had to start and preferably finish with a face-to-face session, and if possible have face-to-face activity in the middle. I now know that if the course design is right, and if the learners are suitably experienced – both big ifs - this is not the case.

Q Anything else you'd like to add?

The impact of “always on” wireless connectivity on learning/online learning will be bigger than many people realise. Partly because of how access devices will change (getting smaller, more multipurpose, and in some respects less usable), and partly because of how different kinds of data will be available to be integrated into the content (for example positional, location-specific, or “friends-close-by” data).

Hope you found the questions stimulating. Thanks for your answers.

The productivity puzzle: bad news for the "education, education, education" crowd

Here is an eye-catching section from Ricardo Hausmann's "The Education Myth":

And there is more bad news for the “education, education, education” crowd: Most of the skills that a labor force possesses were acquired on the job. What a society knows how to do is known mainly in its firms, not in its schools. At most modern firms, fewer than 15% of the positions are open for entry-level workers, meaning that employers demand something that the education system cannot – and is not expected – to provide.

When presented with these facts, education enthusiasts often argue that education is a necessary but not a sufficient condition for growth. But in that case, investment in education is unlikely to deliver much if the other conditions are missing. After all, though the typical country with ten years of schooling had a per capita income of $30,000 in 2010, per capita income in Albania, Armenia, and Sri Lanka, which have achieved that level of schooling, was less than $5,000. Whatever is preventing these countries from becoming richer, it is not lack of education.

A country’s income is the sum of the output produced by each worker. To increase income, we need to increase worker productivity. Evidently, “something in the water,” other than education, makes people much more productive in some places than in others. A successful growth strategy needs to figure out what this is.

Make no mistake: education presumably does raise productivity. But to say that education is your growth strategy means that you are giving up on everyone that has already gone through the school system – most people over 18, and almost all over 25. It is a strategy that ignores the potential that is in 100% of today’s labor force, 98% of next year’s, and a huge number of people who will be around for the next half-century. An education-only strategy is bound to make all of them regret having been born too soon.

Hausmann is former Chief Economist of the Inter-American Development Bank, and Professor of the Practice of Economic Development at Harvard University.

 

Prevent death on the roads by better treatment of obstructive sleep apnoea

Edited image of crash site from BBC News

In August 2006 my nephew Toby was killed by a truck whose driver, Colin Wrighton, an obstructive sleep apnoea sufferer, had "blacked out" (Wrighton's phrase) at the wheel. The crash-scene is above.

I wrote in Fortnightly Mailing about the issue in the years following the CPS decision not to prosecute Wrighton. Examples:

• March 2008 - Joined up government needed to prevent road deaths (after we submitted written evidence about sleep apnoea to Parliament's Transport Committee);
• August 2008 - Coroner calls for Government action on sleep apnoea (after representing my family at Toby's inquest);
• May 2009 - Finding and treating lorry drivers with sleep apnoea (after my MP Meg Munn gave an outstanding speech about sleep apnoea in an adjournment debate in the House of Commons).

Last week we launched the Four-week wait campaign for the treatment of obstructive sleep apnoea syndrome [22p PDF], an attempt to get renewed attention in the UK on the dangers of untreated obstructive sleep apnoea, and, in particular, to guarantee that vocational drivers can be treated for the condition within four weeks, thereby limiting or entirely eliminating the need for them to surrender their license and their livelihood following a diagnosis.

Today, Radio 4's iPM programme broadcast a 15 minute piece about OSA, featuring interviews from 2008 with my sister and brother in law (Toby's parents), and with Colin Wrighton; and a new, long and informative interview with Professor John Stradling, a sleep specialist who is closely involved in the four week wait campaign. Here is a recording of the interview [20MB MP3 file - you may need to "right click" and save the file locally in order to play it]. Or you should be able to stream it from the BBC's web site. I've also uploaded a five-page text transcript of the programme [23kB PDF].

If you want to help us achieve our objectives, write to your MP urging him or her to press the Department of Health, NICE, DVLA, and HSE to work together to ensure that vocational drivers diagnosed with obstructive sleep apnoea are given a cast iron guarantee to be treated within less than 4 weeks. Take particular note of the BBC's interview with John Stradling, when he talks about:

• the proportion of heavy goods vehicle drivers with OSA (~15%);
• the proportion of road accidents due to excessive daytime sleepiness (~20%);
• the extent to which, when an articulated lorry jackknifes, this is normally the consequence of the driver nodding off.

Thank you. 

Brewster Kahle's Internet Archive matters more and more

Here are some links to a couple of past posts about the Internet Archive, which is playing a crucial role in archiving the Web and digital artefacts such as films, games and software:

• 13/3/2009 - Brewster Kahle - the man behind the Wayback Machine;
• 8/10/2012 - Brewster Kahle and the love of books - physical and digital.

Two recent pieces give added emphasis to the importance of Kahle's work, from different angles:

• 26/1/2015 - The Cobweb: Can the Internet be archived? by Jill Lepore in the New Yorker;
• 28/1/2015 - As Google abandons its past, Internet archivists step in to save our collective memory, by Andy Baio.

John Hattie interviewed for Radio 4 by Sarah Montague

Snipped from www.bbc.co.uk/programmes/b04dmxwl

I don't know how long the BBC web-site will carry this 28 minute interview with John Hattie, author of Visible Learning, and respected authority on school-effectiveness.

But in case you cannot access the interview, here is a professionally created transcript [12 page 40kB PDF - not sure what is any IPR issues pertain to this.....].

Taken in the round, Hattie provides a calm and witty counter to many of the ideas used by what Pasi Sahlberg memorably describes as the Global Educational Reform Movement. [See also this 2012 interview with Pasi Sahlberg by John Hattie.]

 

Citizen Maths - an open online level 2 maths course

Over the last 18 months, more-or-less since getting stretchered off a mountain in Norway, I've been leading work funded by the Ufi Charitable Trust to create Citizen Maths which is an open online maths course for adults.

In contrast to many (most?) open online courses the course is at what in England is known as "Level 2", which is the level that 16-year olds are expected to achieve.

The Citizen Maths web site went live on Wednesday, since when people have been signing up on the course, which is built in/on the cloud-based Google Course Builder. If adult education, the learning or teaching of mathematics, or online learning interest you, have a look at Citizen Maths, which tries to put into practice some of the things I've learned about open online courses since doing Peter Norvig and Sebastian Thrun's AI MOOC three years ago. 

You may also be interested in some of the supporting material on the Citizen Maths "Information Hub", for example this piece written with Dave Pratt [also this BBC report] about the thinking behind Citizen Maths, or these informal reflections by me and Dick Moore from an excellent 2-day workshop organised in June by Google in Zurich.

As an aside, my involvement in Citizen Maths, and in the creation of another and very different open online course, seem to have interfered with my ability to write Fortnightly (sic) Mailing. I don't fully understand why this is, and it is only partly explained by the "easy-way-out-that-is-Twitter" - see @sebschmoller and, latterly, @citizenmaths.

I think the probable reason is that if you are responsible for building something when the stakes are quite high, and when you are working with partners (true in both projects), it does not feel quite right to be public about how things are going or about what you are learning from the work. Or, perhaps it's the case that the additional care needed in how you frame things, decide what would and what would not be prudent or fair to say etc., makes writing "too complicated".

And the connection with being stretchered off? There is one. But it's complicated.