Sunday, January 21, 2018

Another 20 million faces

Just over a year ago, Microsoft launched some software that would guess how old you were. Millions of people were persuaded to donate a selfie to Microsoft in return for playing this game. See my post 85 Million Faces (October 2016).

Google's latest face-collecting gimmick is to find a painting that looks like you. Although the Arts and Culture app was originally launched in 2015, the face-matching feature was only added last month. This weekend the app shot to the number one slot in the downloads chart, and 20 million selfies (and counting) have already been donated to Google.

As @ArwaM comments, facial recognition technology allows Google to find the artwork you most resemble – but it also supports the rise of the surveillance state.

And yet Google cannot (yet) compete with old-fashioned serendipity. Before Museum-Doppelgänger-Hunt was an app, it was a viral meme, featuring (among others) @fleezee.

But there have been other Doppelgänger-Hunts before, using Face Recognition software. For example, the TwinStrangers project. So which is the egg and which the chicken?

Rebecca Fleenor, I'm on the front page of Reddit. This is how it feels (CNET, 13 September 2017)

Christine Hauser, Meet your art twin: a 400-year-old with an oily complexion (New York Times, 17 Jan 2018)

Arwa Mahdawi, Finding your museum doppelganger is fun – but the science behind it is scary (Guardian, 16 January 2018)

Rosie Spinks, Why the Art Museum Doppelgänger meme is to profoundly addictive (Quartzy, 2 January 2018)

Der fremde Zwilling (Spiegel, 15 April 2015) in German

Monday, January 15, 2018

Bus Safety Announcement

Transport for London (TfL) reckons around 3000 people are injured every year by slips, trips and falls on London buses. So it is running trials of an automated system that announces the departure of the bus from the stop.
"Please hold on, the bus is about to move"

or as Bon Jovi might say
"We've gotta hold on ready or not."

The problem is that these alerts often come after the bus is already halfway down the road.
"Whoa, we're half-way there."

As the BBC News explains, the timing of the alert is based on the average amount of time a bus would spend at a bus stop, and is often hopelessly inaccurate. Passengers have taken to social media in droves to complain or mock. Many have wondered whether it was such a problem in the first place, and whether an alert would help to alleviate the problem. Others have pointed out the potential value of such an alert for certain categories of passenger - such as the elderly or visually impaired - but of course this only works if the alert comes at the right time.

I haven't spoken to anyone at TfL about this, but I can imagine what happened. In order to get a trial up and running quickly, they didn't have time (or permission) to link the alert with any of the systems on board the bus that could have sent a more accurate event signal. So we have a stand-alone system, knocked up quickly, as an experimental solution to a problem that most people hadn't previously recognized. In the trimodal scheme, this is a classic Pioneer project.
"For love we'll give it a shot."

So if the trial isn't laughed into touch, then maybe the Settlers can take over and do the alert properly.
"Take my hand, we'll make it. I swear."

And the Town Planners can come up with a joined-up long-term vision for passenger comfort and safety. Altogether now ...
"Whoa, livin' on a prayer."

Londoners hit out at 'mistimed' bus safety alerts (BBC News, 14 January 2018)

Nadia Khomami, Please hold on: TfL urged to get a grip over annoying bus warnings (Guardian, 15 January 2018)

Londoners baffled by 'bonkers' bus safety announcements warning them 'the bus is about to move' (Evening Standard,15 January 2018).

For more on Trimodal IT, see my post Beyond Bimodal (May 2016)

Wednesday, December 27, 2017

Automated Tetris

Following complaints that Amazon sometimes uses excessively large boxes for packing small items, the following claim appeared on Reddit.

"Amazon uses a complicated software system to determine the box size that should be used based on what else is going in the same truck and the exact size of the cargo bay. It is playing automated Tetris with the packages. Sometimes it will select a larger box because there is nothing else that needs to go out on that specific truck, and by making it bigger, it is using up the remaining space so items don't slide around and break. This actually minimizes waste and is on the whole a greener system. Even if for some individual item it looks weird. It's optimizing for the whole, not the individual." [source: Reddit via @alexsavinme]

Attached to the claim is a link to @willknight's 2015 article about Amazon's robotic warehouses. The article mentions the packing problem but doesn't mention the variation of box sizes.

The claim quickly led to vigorous debate, both on Reddit and on Twitter. Here are a selection of the argument and counter-arguments.

  • Suggesting that the Reddit claim was based on a misreading of the MIT article.
  • Asserting that people working in warehouses (Amazon and other) were unaware of such an algorithm. (As if this were relevant evidence.)
    • Evidence that equally sophisticated algorithms are in use at other retailers and logistics companies. (Together with an assumption that if others have them, Amazon must definitely have them.)
    • Evidence that some operational inefficiencies exist at Amazon and elsewhere. (What, isn't Amazon perfectly optimized yet?)
      • Providing evidence that computer systems would not always recommend the smallest possible box. For example, this comment: "At Target the systems would suggest a size but we could literally use whatever we wanted to. I constantly put stuff in smaller boxes because it just made so much more sense." (Furthermore, the humans being able to frustrate the intentions of the software.)
      • Suggesting that errors in box sizes are sometimes caused by mix-up of units - one item going in a box large enough for a dozen.
      • Pointing out that the solution described above would only work for transport between warehouses (where the vehicle is full for the whole trip) but wouldn't work for "last mile" delivery runs (where the vehicle becomes progressively more empty during the trip).
      • Pointing out that the "last mile" is the most inefficient part of the journey. (But this doesn't stop retailers looking for efficiency savings earlier in the journey.)
      • Pointing out that there were more efficient solutions for preventing packages shifting in transit - for example, inflatable bags.
      • Pointing out that an overlarge box merely displaces the problem - the item can be damaged by sliding around inside the box.
      • Complaining about the ethics, employment policies and environmental awareness of Amazon.
      • Denigrating the intelligence and diligence of the workers in the Amazon warehouse. (Lazy? Really?)

      Some people have complained that as the claim is evidently false, it counts as fake news and should be deleted. But it is certainly true that retailers and logistics companies are constantly thinking about ways of reducing packaging and waste, and there are several interesting contributions to the debate, even if some of the details may not quite work.

      It's also worth noting that the claim is written in a highly plausible style - that's just how people in that world would talk. So maybe someone has come across a proposal or pilot or patent application along these lines, even if this exact solution was never fully implemented.

      Some may doubt that such a solution would be "greener on the whole". But any solution architect should get the principle of "optimizing for the whole, not the individual". (Not always so easy in practice, though.) 

      Will Knight, Inside Amazon’s Warehouse, Human-Robot Symbiosis (MIT Technology Review, 7 July 2015)

      Wikipedia: Packing Problems

      Thursday, December 14, 2017

      Expert Systems

      Is there a fundamental flaw in AI implementation, as @jrossCISR suggests in her latest article for Sloan Management Review? She and her colleagues have been studying how companies insert value-adding AI algorithms into their processes. A critical success factor for the effective use of AI algorithms (or what we used to call expert systems) is the ability to partner smart machines with smart people, and this calls for changes in working practices and human skills.

      As an example of helping people to use probabilistic output to guide business actions, Ross uses the example of smart recruitment.
      But what’s the next step when a recruiter learns from an AI application that a job candidate has a 50% likelihood of being a good fit for a particular opening?

      Let's unpack this. The AI application indicates that at this point in the process, given the information we currently have about the candidate, we have a low confidence in predicting the performance of this candidate on the job. Unless we just toss a coin and hope for the best, obviously the next step is to try and obtain more information and insight about the candidate.

      But which information is most relevant? An AI application (guided by expert recruiters) should be able to identify the most efficient path to reaching the desired level of confidence. What are the main reasons for our uncertainty about this candidate, and what extra information would make the most difference?

      Simplistic decision support assumes you only have one shot at making a decision. The expert system makes a prognostication, and then the human accepts or overrules its advice.

      But in the real world, decision-making is often a more extended process. So the recruiter should be able to ask the AI application some follow-up questions. What if we bring the candidate in for another interview? What if we run some aptitude tests? How much difference would each of these options make to our confidence level?

      When recruiting people for a given job, it is not just that the recruiters don't know enough about the candidate, they also may not have much detail about the requirements of the job. Exactly what challenges will the successful candidate face, and how will they interact with the rest of the team? So instead of shortlisting the candidates that score most highly on a given set of measures, it may be more helpful to shortlist candidates with a range of different strengths and weaknesses, as this will allow interviewers to creatively imagine how each will perform. So there are a lot more probabilistic calculations we could get the algorithms to perform, if we can feed enough historical data into the machine learning hopper.

      Ross sees the true value of machine learning applications to be augmenting intelligence - helping people accomplish something. This means an effective collaboration between one or more people and one or more algorithms. Or what I call organizational intelligence.

      Postscript (18 December 2017)

      In his comment on Twitter, @AidanWard3 extends the analysis to multiple stakeholders.
      This broader view brings some of the ethical issues into focus, including asymmetric information and algorithmic transparency

      Jeanne Ross, The Fundamental Flaw in AI Implementation (Sloan Management Review, 14 July 2017)

      Saturday, December 02, 2017

      The Smell of Data

      Retailers have long used fragrances to affect the customer in-store experience. See for example Air/Aroma.

      So perhaps we can use smell to alert consumers to dodgy websites? An artist and graphic designer, Leanne Wijnsma, has built what is basically an air-defreshener: a hexagonal resin block with a perfume reservoir inside, which connects over Wi-Fi to your computer. When it notices a possible data leak (like the user connecting to an unsecured Wi-Fi network, or browsing a webpage over an unsecure connection) — puff! It releases the smell of data.

      James Vincent, What does a data leak smell like? This little device lets you find out (Verge, 31 Aug 2017)

      That's all very well, but it only sniffs out the most obvious risks. If you want to smell the actual data leak, you'd need a device that released a data leak fragrance when (or perhaps I should say whenever) your employer or favourite online retailer is hacked. Or maybe a device that sniffed around a corporate website looking for vulnerabilities ...

      I'm sure my regular readers don't need me to spell out the flaws in that idea.

      Related posts

      Pax Technica - On Risk and Security (November 2017)
      UK Retail Data Breaches (December 2017)

      UK Retail Data Breaches

      Some people talk as if data protection and security must be fixed before May 2018 because of GDPR. Wrong. Data protection and security must be fixed now.

      Morrisons (2014)

      The High Court has just found Morrisons to be liable for a leak of employee data by a disaffected employee in 2014. (The perpetrator got eight years in jail.)

      Sports Direct (2016)

      A hacker obtained employee details in September 2016, but Sports Direct failed to communicate the breach to the affected employees.

      CEX (2017)

      Second-hand gadget and video games retailer Cex has said up to two million customers have had their data stolen in an online breach

      Zomato (2017)

      Up to 17 million users affected by data breach at restaurant search platform Zomato

      Tesco Bank (2016)

      Cyber thieves steal £2.5m

      Related posts

      The Smell of Data (December 2017)

      Monday, September 25, 2017

      Regulating Platforms

      On Friday, Transport for London (TfL) declared that Uber was not fit and proper to hold a private hire operator licence. Uber's current licence expires next week. However, Uber can continue to operate in London until any appeal processes have been exhausted. (TFL Press Release, 22 September 2017)

      By Saturday afternoon, a petition in Uber's favour had raised half a million signatures. Uber seems to put more energy into campaigning against evil regulators than into operating within the regulations, and was evidently already prepared for this fight. (You don't send out messages to millions of customers at the drop of a hat without a bit of forward planning.) As Emine Saner writes,
      "Calling for better legislation certainly is not as exciting as a glossy app, or whipped-up social media reaction, but it may make your trip home safer – and would be a better use of online petitions."

      The protests follow a number of well-worn arguments
      • Many users of the Uber service (especially young women) have become dependent on a cheap, convenient and supposedly safer alternative to public transport and expensive taxis.
      • Many drivers have borrowed heavily to invest in the Uber business model, and fear being thrown into penury.
      • This is an anti-competitive and technologically backward move, prompted by entrenched interests. And as TfL is itself a transport operator, it is not appropriate that TfL should regulate its competitors.

      None of these arguments can be taken completely at face value.

      • It is true that many women believe the Uber model is safer than the alternatives; however, some women have been raped, and other women have had extremely scary experiences. Uber is accused of failing to carry out proper checks, and failing to report serious incidents.
      • Uber service is cheap not only because it cuts costs and exploits its drivers, but also because it is subsidized by Uber investors. This looks suspiciously like predatory pricing rather than fair competition. Analysts such as Izabella Kaminska argue that Uber will only become profitable when it has driven its competitors out of business, at which point it will be able to increase its prices. Like much of Silicon Valley, it appears to operate according to the Peter Thiel anti-competition playbook. Even Steve Bannon has been heard arguing for closer regulation of what are effectively monopoly platforms.
      • Technology companies such as Uber sometimes describe themselves as "disruptive". While it is true that disruptions sometimes yield socioeconomic benefits, the belief that disruption is always good for competition is based on ideology rather than evidence. Regulation is generally opposed to disruption.
      • And as Stephen Bush points out, it's not as a digital start-up company that Uber has fallen foul of regulations, but as an old fashioned minicab operator. (As John Bull explains, Uber London is just a minicab company; the app is operated by Uber BV in the Netherlands. This corporate separation helps Uber to finesse both regulation and tax.) Persuading politicians and economists to see Uber as a shining example of technological progress is just "a very, very clever marketing trick".

      I'm quoting Steve Bannon because I'm just amazed to find something I agree with him about.  Regulating platforms is not the same as regulating regular companies, and the general art of regulation needs a kick up the proverbial. However, that is no reason to diss the current regulations or regulators, who are doing the best they can with insufficient regulatory mechanisms and resources. Experience from other cities shows that if Uber can't get its act together, there are plenty others that can.

      John Bull, Understanding Uber: It’s Not About The App (Reconnections 25 September 2017)

      Stephen Bush, The right are defending Uber, because they don't really understand it (New Statesman 22 September 2017)

      Martin Farrer, Nadia Khomami et al, More than 500,000 sign petition to save Uber as firm fights London ban (Guardian 23 September 2017)

      Ryan Grim, Steve Bannon Wants Facebook and Google Regulated Like Utilities (The Intercept, 27 July 2017)

      Hubert Horan, Will the Growth of Uber Increase Economic Welfare? (September 14, 2017)

      Izabella Kaminska. For references see earlier post Uber Mathematics 2 (December 2016)

      Sam Levine,'There is life after Uber': what happens when cities ban the service? (Guardian 23 September 2017)

      Jason Murugesu, Night bus or black cab - what will save stranded Londoners post-Uber? (New Statesman 22 September 2017)

      Andrew Orlowski, Why Uber isn't the poster child for capitalism you wanted (The Register, 26 September 2017)

      Emine Saner, Will the end of Uber in London make women more or less safe? (Guardian, 25 September 2017)

      Related posts (with further references): Platform, Regulation, Uber

      Wednesday, August 16, 2017

      Digital Disruption, Delivery and Differentiation in Fast Food

      What are the differentiating forces in the fast food sector? Stuart Lauchlan hears some contrasting opinions from a couple of industry leaders.

      In the short term, those fast food outlets that offer digital experience and delivery may get some degree of competitive advantage by reaching more customers, with greater convenience. Denny Marie Post, CEO at Red Robin Gourmet Burgers, sees the expansion of third-party delivery services as a strategic priority. So from agility to reach.

      But Lenny Comma, CEO of Jack in the Box, argues that this advantage will be short-lived. Longer-term competitive advantage will depend on the quality of the brand. So from assurance to richness.

      Stuart Lauchlan, Digital and delivery – which ‘D’ matters most to the fast food industry? Two contrasting views (Diginomica, 16 August 2017)

      Related post: Reach, Richness, Agility and Assurance (Aug 2017)

      Tuesday, August 15, 2017

      Reach, Richness, Agility and Assurance

      The concept of TotalData™ implements the four dimensions of data and information - reach, richness, assurance and agility. But where did these dimensions come from?

      I first encountered these four dimensions in discussions of net-centricity, which spilled out from the US defence world into the commercial world over ten years ago. Trying to dig up the original material recently, I found a military version in a report written in 2005 by the Association for Enterprise Integration (AFEI) for the Net-Centric Operations Industry Forum (NCOIF).

      Going further back, the first two dimensions - reach and richness - had been discussed by Evans and Wurster before the turn of the millennium. They argued that old technologies had forced you to choose (either/or) between reach and richness, whereas the new technologies emerging at that time allowed you to have both/and.

      Source: Evans and Wurster 1997

      The authors also introduced the concept of affiliation, by which they meant transparency of relationships - for example, knowing whether the intermediary agent is working for you or working for the other side. Or both. And knowing who really wrote all those "customer reviews".

      According to the authors, it would be these three factors - reach, richness and affiliation - that would determine the success of e-commerce. Clearly some sectors would be more open to these factors than others - according to The Economist in February 2000, online trade was then dominated by business-to-business (B2B). The three factors identified some of the challenges facing other sectors, including professional services, in going online. As Duncan, Barton and McKellar argued for legal firms, "The Web provides Reach, but offering Richness and the sense of community required for creating and sustaining relationships with visitors could be difficult."

      Meanwhile, new architectural thinking had shown ways of resolving the traditional trade-off between speed (agility) and quality (assurance). (A very early version of this was known as Bimodal IT. Some industry analysts are still pushing this idea.)

      When agility and assurance were added to reach and richness to produce the four dimensions of net-centricity, affiliation appears to have been divided between community (reach) and trust (assurance). But the importance of affiliation was never entirely forgotten. As Commander Chakraborty observes, "organisational affiliations and culture ... play very significant roles in a networked environment."

      So whatever happened to net-centricity? It has been replaced by data-centricity, which, as Dan Risacher argues, is probably a more accurate term anyway. Or as we call it at Reply, TotalData™.

      Notes and References

      Much of the original material for the NCOW Reference Model is no longer available. This includes the pages referenced from Wikipedia: NCOW (retrieved 8 August 2017). Net-centric concepts were incorporated into DODAF Version 1.5 (April 2007).

      Define and Sell (Economist, 24 Feb 2000)

      AFEI, Industry Best Practices in Achieving Service Oriented Architecture (SOA) (NCOIF, April 2005)

      Devbrat Chakraborty, Net-Centricity to Ne(x)t-Centricity (SP's Navel Forces, Issue 4/2011)

      Peter Duncan, Karen Barton and Patricia McKellar, Reach and Rich: the new economics of information and the provision of on-line legal services in the U.K. (16th Bileta Annual Conference, 2001)

      Philip Evans and Thomas S. Wurster, Strategy and the New Economics of Information (Harvard Business Review, Sept-Oct 1997)

      Philip Evans and Thomas Wurster, Blown to Bits - How the New Economics of Information Transforms Strategy (Boston Consulting Group, 2000) - excerpts. See also reviews by McRae and O'Keefe.

      Hamish McRae, The business world: Three factors that lead to successful e-commerce (Independent, 17 November 1999) - review of Evans and Wurster (2000)

      Jordan Moskowitz, Richness versus Reach (Service Channel, 29 Jan 2013)

      Terry O'Keefe, The strategy of information: Richness and reach (Atlanta Business Journal, 1 November 1999) - review of Evans and Wurster (2000)

      Dan Risacher, The Fundamentals of Net-Centricity (a little late) (4 February 2013)

      Related Posts: Beyond Bimodal (May 2016), New White Paper - TotalData™ (August 2016)

      TotalData™ is a registered trademark of Reply Ltd.

      Tuesday, July 18, 2017

      On the Nature of Platforms

      There are several ways of thinking about platforms.

      Economists tend to view platforms as essentially containers for transactions. Canonical examples: Amazon, Airbnb, iTunes, Netflix, Uber.

      One of the economic advantages of these transaction platforms is that they also act as container for content. When it launched in 1995, the Amazon website boasted a million books - far more than you could find in any bookshop. (This is related to the concept of the Long Tail.) So it becomes a place you can browse books and check reviews, independently of any intention to buy.

      Transaction platforms may also enable a significant reduction in transaction costs. This creates an opening for micro-transactions of various kinds - in other words transactions that would previously have been too small to be economically viable. Smaller-grained transactions can allow previously under-utilized assets to be used more economically - for example selling an empty passenger seat on a car journey.

      Transaction platforms may also act as a container for data and/or metadata. Dave Chaffey describes "Customers Who Bought X ... Also Bought Y" as Amazon's signature feature. ( case study, 30 June 2014)

      This notion of platform can be extended to containers of other modes of activity or collaboration or exchange, where there may be no direct financial transaction. Canonical examples: Facebook, PlayStation Network, Linked-In, Skype.

      There are various business models underlying these activity platforms, including freemium (Linked-In, Skype), post-sale delivery and engagement (PSN), and advertising. As many people have observed, Facebook inherits a principle that was originally formulated for commercial television - if you are not paying, you are the product. In other words, the underlying transaction is the one between Facebook and its advertisers, whereby Facebook rents out the user to its paying customers.

      These platforms are typically described as two-sided or multi-sided. Among other things, multi-sidedness implies some choices about pricing strategy - how to distribute the costs and added-value of the platform between the different sides. For example, credit card provides a transaction platform between consumers and merchants - the credit card company has a choice whether to charge everything to the merchants or to charge the consumers as well. And Facebook and Google provide user services for free, although perhaps one day we shall be so addicted to their services that they can make us pay hard cash to continue.

      A different way of thinking about platforms is as a container for capabilities or services. Here, the canonical example would be Amazon Web Services (AWS). At the CBDI Forum, we were writing about AWS over ten years ago, but many people only became aware of AWS when it grew into a massive business in its own right. (Amazon and eBay, August 2004)

      The key idea here is that you can build a business on top of a platform. Thus a start-up online retailer doesn't need to build all the necessary capabilities in-house, because there is a platform of services already available. In the 1990s, telecoms companies were looking for ways to create value-added services on top of the basic communication platforms.
      Many companies already have a platform, but they are trying to raise it. For example, the traditional role for telecoms companies is as a platform of telecoms connectivity. But it has been obvious for ages that there is no long-term profitability for telecoms from providing services at this level. So telecoms companies have long understood the need to raise the platform, to offer higher-value services. But they are still struggling to formulate and implement this strategic change. Why is it so difficult? (Business as a Platform, March 2006) 

      Similar structures can be found in the physical world. In addition to managing real platforms at railway stations, Network Rail provides a "platform" on which the train operating companies can run their business. In theory these are services with strict service level agreements and contractual or regulatory penalties, although the actual stack geometry is arguably flawed. (Business Service Architecture - Railway Edition, June 2006)

      In retail, some large department stores have turned themselves into marketplaces in which franchise retailers can sell their products. Other retailers have experimented with a business model in which the goods are owned by the supplier up to the point at which they are purchased by the supplier. Thus the store becomes a platform for the supplier to merchandise and sell products. See also Nick Vitalari, Walmart and The Power of the Business Platform (Sept 2011).

      Platforms are sometimes described as more or less open or closed. For example, the Open Banking Platform. Platform controllers often seek to impose quality or technical constraints on businesses using the platform - for example, Apple iTunes. Thus the notion of openness has a range of meanings, from market openness (e.g. no barriers to entry and exit) to technological openness (e.g. flexibility of mechanism). (Types of Openness, November 2001)

      So when business strategy consultants talk about a platform business, this can also refer to the flexible and open-ended exploitation of an asset or capability, to create or co-create value in as many ways as possible. For example, here is John Hagel in 2006, talking about Steve Jobs and Disney.
      In a world of scarce attention, creators of media products will need to compete with those who re-conceive media products as platforms. What is the difference? Products are designed to be used on a standalone basis – you buy it and you view it or listen to it in the specific way the content creator intended. Platforms are designed to be built upon – they create opportunities for the original creator, third parties or the customers themselves to extend, enhance and tailor the content in ways that the original creator never anticipated. Offered as a platform, content can create far more value than any equivalent standalone product. (Disney, Pixar and Jobs, Feb 2006)

      Finally, we may note that although many of these platforms may be described as "digital", many of the same basic characteristics can be found in both digital and non-digital modes. And what even counts as "non-digital" these days, when every aspect of our lives can be wired to the Internet? So I prefer not to talk about digital platforms any more - they are just platforms.

      Further Reading

      Philip Boxer, What Distinguishes a Platform Strategy? (Asymmetric Design, May 2012)

      Diane Coyle, The Social Life of Platforms (Enlightenment Economics, May 2016)

      For John Hegel's latest thinking about platforms, see The Big Shift in Business Platform Models (Edge Perspectives, January 2017)

      Related posts: Multi-Sided Platform Strategies (April 2013)