Sunday, April 10, 2022

Lie Detectors at Airports

@jamesvgingerich reports that the EU is putting lie detector robots on its borders. @Abebab is horrified.


There are several things worth noting here.

Firstly, lie detectors work by detecting involuntary actions (eye movements, heart rate) that are thought to be a proxy for mendacity. But there are often alternative explanations for these actions, and so the interpretation of these is highly problematic. See my post on Memory and the Law (June 2008)

Secondly, there is a lot of expensive and time-wasting technology installed at airports already, which has dubious value in detecting genuine threats, but may help to make people feel safer. Bruce Schneier calls this Security Theatre. See my posts on the False Sense of Security (June 2019) and Identity, Privacy and Security at Heathrow Terminal Five (March 2008).

What is more important is to consider the consequences of adding this component (whether reliable or otherwise) to the larger system. In my post Listening for Trouble (June 2019), I discussed the use of Aggression Detection microphones in US schools, following an independent study that was carried out with active collaboration from the supplier of the equipment. Obviously this kind of evaluation requires some degree of transparency.

Most important of all is the ethical question. Is this technology biased against certain categories of subject, and what are the real-world consequences of being falsely identified by this system? Is having a human in the loop sufficient protection against the dangers of algorithmic profiling? See Algorithmic Bias (March 2021).

Given the inaccuracy of detection, there may be a significant rate of false positives and false negatives. False positives affect the individual concerned, suffering consequences ranging from inconvenience and delay to much worse. False negatives mean that a person has got away with an undetected lie, so this has consequences for society as a whole. 

How much you think this matters depends on what you think they might be lying about, and how important this is. For example, it may be quicker to say you packed your suitcase yourself and it hasn't been out of your sight, even if this is not strictly true, because any other answer may trigger loads of other time-wasting questions. However, other lies may be more dangerous ...


For more details on the background of this initiative, see

Matthias Monroy, EU project iBorderCtrl: Is the lie detector coming or not? (26 April 2021)

Thursday, December 23, 2021

The Sad Reality of Chatbotics

As I noted in my previous post on chatbotics Towards Chatbot Ethics (May 2019), the chatbot has sometimes been pitched as a kind of Holy Grail. Which prompts the question I discussed before - whom shall the chatbot serve?

Chatbots are designed to serve their master - and this is generally the organization that runs them, not necessarily the consumer, even if you have paid good money to have one of these curious cylinders in your home. For example, Amazon's Alexa is supposed to encourage consumers to access other Amazon services, including retail and entertainment -  and this is how Amazon expects to make a financial return on the sale and distribution of these devices.

But how well do they work even for them? The journalist Priya Anand (who tweets at @PriyasIdeas) has been following this question for a while. Back in 2018, she talked to digital marketing experts who warned that voice shopping was unlikely to take off quickly. Her latest article notes the attempts by Amazon Alexa to nudge consumers into shopping, which may simply cause some frustrated consumers to switch the thing off altogether. Does this explain the apparently high attrition rates?

If you are selling a device at a profit, it may not matter if people don't use it much. But if you are selling a device at an initial loss, expecting to recoup the money when the device is used, then you have to find ways of getting people to use the thing. 

Perhaps if Amazon can use its Machine Learning chops to guess what we want before we've even said anything, then the chatbots can cut out some of the annoying chatter. Apparently Alexa engineers think this would be more natural. Others might argue Natural's Not In It. (Coercion of the senses? We're not so gullible.)

Priya Anand, The Reality Behind Voice Shopping Hype (The Information, 6 August 2018)

Priya Anand, Amazon’s Alexa Stalled With Users as Interest Faded, Documents Show (Bloomberg, 22 December 2021)

Daphne Leprince-Ringuet, Alexa can now guess what you want before you even ask for it (ZDNet, 13 November 2020)

Tom McKay, Report: Almost Nobody Is Using Amazon's Alexa to Actually Buy Stuff (Gizmodo, 6 August 2018)

Chris Matyszczyk, Amazon wants you to keep quiet, for a brilliantly sinister reason (ZDNet, 4 November 2021)

Related posts: Towards Chatbot Ethics (May 2019), Technology and the Discreet Cough (September 2019)

Tuesday, September 28, 2021

Automation and the Red Queen Effect

Product vendors and technology advisory firms often talk about accelerating automation. A popular way of presenting advice is in the form of an executive survey - look, all your peers are thinking about this, so you'd better spend some money with us too. Thus the Guardian reports a survey carried out by one of the large consulting firms, which concluded that almost half of company bosses in 45 countries are speeding up plans to automate their businesses. In the early days of the COVID-19 pandemic, many saw this as a great opportunity to push automation further, although more recent commentators have been more sceptical.

Politicians have also bought into this narrative. For example, Barack Obama's farewell presidential address referred to the relentless pace of automation.

For technical change more generally, there is a common belief that things are constantly getting faster. In my previous posts on what is sometimes known as the Red Queen Effect, I have expressed the view that perceptions of technological change often seem to be distorted by proximity and a subjective notion of significance - certain types of recent innovation being regarded as more important and exciting than other or older innovations.

Aaron Benanav takes a similar view.

Our collective sense that the pace of labor-saving technological change is accelerating is an illusion. It’s like the feeling you get when looking out of the window of a train car as it slows down at a station: passing cars on the other side of the tracks appear to speed up. Labor-saving technical change appears to be happening at a faster pace than before only when viewed from across the tracks – that is, from the standpoint of our ever more slow-growing economies. Benanav 2020

Benanav also notes that the automation narrative has been around since the days of Karl Marx.

Visions of automated factories then appeared again in the 1930s, 1950s and 1980s, before their re-emergence in the 2010s. Benanav 2019
Meanwhile, Judy Wajcman argues (referencing Lucy Suchman) that the automation narrative typically relies on overlooking the human labour that is required to keep the computers and robots working efficiently - especially the low-paid work of data coders and content checkers. Further evidence of this has recently been published by Phil Jones.


Bosses speed up automation as virus keeps workers home (Guardian, 30 March 2020)

Is the pandemic accelerating automation? Don’t be so sure (Economist, 19 June 2021) subscription required

Aaron Benanav, Automation and the Future of Work. Part One (New Left Review 119 Sept/Oct 2019) Part Two (New Left Review 120, Nov/Dec 2019)

Aaron Benanav, Automation isn't wiping out jobs. It's that our engine of growth is winding down (Guardian, 23 January 2020)

Phil Jones, Work without the worker - Labour in the age of platform capitalism (Verso, 2021) Extract published as Refugees help power machine learning advances at Microsoft, Facebook, and Amazon (Rest of World, 22 September 2021)

Toby McClean, Automation Is Accelerating At The Edge To Improve Workplace Safety, Productivity (Forbes, 5 January 2021) 

Judy Wajcman, Automation: is it really different this time? (British Journal of Sociology, 2017)

Chris Wiltz, Grocery Automation Is Accelerating Thanks to the Coronavirus (Grocery News, 16 April 2020)

Tuesday, April 27, 2021

The Allure of the Smart City

The concept of smart city seems to encompass a broad range of sociotechnical initiatives, including community-based healthcare, digital electricity, housing affordability and sustainability, next-generation infrastructure, noise pollution, quality of air and water, robotic furniture, transport and mobility, and urban planning. The smart city is not a technology as such, more like an assemblage of technologies.

Within this mix, there is often a sincere attempt to address some serious social and environmental concerns, such as reducing the city's carbon footprint. However, Professor Rob Kitchen notes a tendency towards greenwashing or even ethics washing.

Kitchen also raises concerns about civic paternalism - city authorities and their tech partners knowing what's best for the citizenry.

On the other hand, John Koetsier makes the point that If-We-Don't-Do-It-The-Chinese-Will. This point was also recently made by Jeremy Fleming in his 2021 Vincent Briscoe lecture. (See my post on the Invisibility of Infrastructure.)

Meanwhile, here is a small and possibly unrepresentative sample of Smart City initiatives in the West that have reached the press recently.

  • Madrid with IBM
  • Portland with Google Sidewalk - cancelled Feb 2021
  • San Jose with Intel - pilot programme
  • Toronto with Google Sidewalk (Quayside) - cancelled May 2020



Daniel Doctoroff, Why we’re no longer pursuing the Quayside project — and what’s next for Sidewalk Labs (Sidewalk Talk, 7 May 2020)

Rob Kitchen, The Ethics of Smart Cities (RTE 27 April 2019)

John Koetsier, 9 Things We Lost When Google Canceled Its Smart Cities Project In Toronto (Forbes, 13 May 2020) 

Ryan Mark and Gregory Anya, Ethics of Using Smart City AI and Big Data: The Case of Four Large European Cities (Orbit, Vol 2/2, 2019)

Juan Pedro Tomás, Smart city case study: San Jose, California (Enterprise IOT Insights, 5 October 2017)

Jane Wakefield, The Google city that has angered Toronto (BBC News, 18 May 2019), Google-linked smart city plan ditched in Portland (BBC News, 23 February 2021)


See also IOT is coming to town (December 2017), On the invisibility of infrastructure (April 2021)


Wednesday, February 03, 2021

Andy Jassy

Most people still think of Amazon primarily as an online retailer, but the elevation of Andy Jassy to take over from Jeff Bezos as CEO provides further evidence for the strategic importance of Amazon Web Services (AWS) within the Amazon group.

AWS was launched in 2002 and relaunched in 2006. In March 2008, Om Malik published an interview with Ray Ozzie, then the Chief Software Architect at Microsoft, which included some positive comments about AWS. By the end of the year, both Google and Microsoft had announced rival cloud computing offerings. As far as I can see, cloud computing first appeared as an Emerging Technology on the Gartner Hype Curve (it's not a cycle) in 2008, reaching the Peak of Inflated Expectations by 2009.

During that period, I was a software industry analyst, calling out Jeff Bezos and Ray Ozzie as two of the most visionary players in the industry. My colleague Lawrence Wilkes wrote a long report on AWS in 2004. (But the hype around cloud computing took off later, and the broader awareness of AWS is comparatively recent, so I'm not convinced that the classic hype curve applies to this topic.)

Alongside the news of Jassy's elevation, today's tech press also reports that Google Cloud is still making massive losses. So much for the Slope of Enlightenment then.


Jasper Jolly, Bezos leaves Amazon in its prime – keeping it that way is the task (The Guardian, 3 February 2021)

Kieren McCarthy, So Jeff Bezos is stepping back from Amazon to play with his space rockets. Who's this Andy Jassy chap? (The Register, 3 February 2021)

Om Malik, GigaOM Interview: Ray Ozzie (GigaOM, 10 March 2008)

Ron Miller, What Andy Jassy’s promotion to Amazon CEO could mean for AWS (TechCrunch, 2 February 2021)

Simon Sharwood, Google's cloud services lost $14.6bn over three years – and CEO Sundar Pichai likes that trajectory (The Register, 3 February 2021)

Lawrence Wilkes, Amazon and eBay Web Services - The new enterprise applications? (CBDI Journal, October 2004) 

Related posts: Jeff Bezos and Ecosystem Thinking (February 2004), Amazon and eBay (August 2004), Internet Service Disruption (November 2005), Ray Ozzie (March 2008), Utility Computing and Profitability (March 2008)

Also Technology Hype Curve (September 2005)

Sunday, December 29, 2019

The Allure of the Smart Home

What exactly is a smart home, and why would I want to live in one?

I don't think the smart home concept is just about having the latest cool technology or containing some smart stuff. And many of the most commonly discussed examples of smart technology in the home seem to be merely modest improvements on earlier technologies, rather than something entirely new.

Let's look at some smart devices you might have in your home. Programmable thermostats have been available for ages, adjusting heating and/or air conditioning to maintain a comfortable temperature at certain times of day. Modern heating systems can now offer separate controls for each room, and be programmed to reduce your total energy consumption: such systems are typically marketed as intelligent systems. So whatever smart technology is doing in this area looks more like useful improvement than radical change.

Or how about remote control functionality. Remote control devices have been around for a long time, especially for couch potatoes who wished to change TV channels without the effort of walking a few feet across the room. Now we have voice-activated controls, for people who can't even be bothered to search under the cushions for the remote control device. Voice activation may be a bit more technologically sophisticated than pushing buttons, and some artificial intelligence may be required to recognize and interpret the voice commands, but it's basically the same need that is being satisfied here.

Or how about a chatbot to answer your questions? In most cases, the answers aren't hard-wired into the device, but are pulled from some source outside the home. So the chatbot is merely a communication device, as if you had a telephone hotline to Stephen Fry only faster and always available, like several million Stephen Fry clones working in parallel around the clock. (You may choose any other  knowledgeable and witty celebrity if you prefer.)

And the idea that having a chatbot device in your home makes the home itself smart is like thinking that having a smartphone in the pocket of your trousers turns them into smart trousers. Or that having Stephen Fry's phone number attached to your fridge door turns it into a smart fridge.

Of course, a smart system may have multiple components - different classes of device. You might install an intelligent security system, using cameras and other devices, to recognize and admit your children and pets, while keeping the home safe from unwanted visitors.

But surely the concept of smart home means more than just having a number of smart parts or subsystems, it implies that the home itself manifests some intelligence at the whole-system level. The primary requirement seems to be that these smart devices are connected, not to the world outside the home, but to each other, enabling them to orchestrate things. Not just home automation, but seamless home automation.

For example, suppose I make my home responsible for getting me to work on time. My home computer could monitor the traffic reports or disruption on public transport, check with my car whether I needed to allow extra time to refuel, send a message to my alarm clock to wake me up at the optimal time, having also instructed the heating system when to switch the boiler on.

Assuming I do not wish my movements to be known in advance by burglars and kidnappers, all of these messages need to be secure against eavesdropping. It isn't obvious to me why it would be necessary to transmit these messages via servers outside my house. Yes I know it's called the internet of things, but does that mean everything has to go via the internet?

Well yes apparently it does, if we follow the recently announced Connected Home over IP (CHIP) standards, to be developed jointly by Amazon, Apple, Google, and most of the other key players in the smart home market.

Many of those who commented on the Register article raised concerns about encryption. It seems unlikely at this point that the tech giants will be keen on end-end encryption, because surely they are going to want to feed your data into the hungry mouths of their machine learning starlings. So whatever security measures are included in the CHIP standards, they will probably represent a compromise, appearing to take security seriously but not seriously impeding the commercial and strategic interests of the vendors. Smart for them, not necessarily for us.

Sometimes it seems that the people who benefit most from the smart home are not those actually living in these homes but service providers, using your data to keep an eye on you. For example, landlords:
Smart home technology is an alluring proposition for the apartment industry: Provide renters with a home that integrates with and responds to their lifestyle, and ­increase rents, save on energy, and collect useful resident population data in return. Kayla Devon
Internet-connected locks and facial recognition systems have raised privacy concerns among tenants across the country. A sales pitch directed at landlords by a smart-home security company indicates that the technology could help them raise rental prices and potentially get people evicted. Alfred Ng
We should pass a law that would hold smart access companies to the highest possible standard while making certain that their technology is safe, secure and reliable for tenants. Michael McKee

Energy companies have been pushing smart meters and other smart technologies, supposedly to help you reduce your energy bills, but also to get involved in other aspects of your life. For example, Constellation promotes the benefits of smart home technology for maintaining the independence of the elderly, while Karen Jorden mentions the possibility of remote surveillance by family members living elsewhere.
Smart technology that recognizes patterns, such as the morning coffee-making routine mentioned earlier, could come in handy when those patterns are broken, perhaps alerting grown children that something may be amiss with an elderly parent. Karen Jordan

As Ashlee Clark Thompson points out, this kind of remote surveillance can benefit the children as well as the parents, providing peace of mind as well as reducing the need for physical visits to check up. 

And doubtless the energy companies have other ideas as well. According to Ross Clark

Scottish and Southern Electricity Networks has proposed a system in which it will be able to turn off certain devices in our homes ... when the supply of electricity is too small to meet demand.

Finally, Ian Dunt grumbled that his smart thermostat was like having a secret flatmate.

and got dozens of Tweets in reply, from people with similar frustrations.

So we keep coming back to the fundamental ethical question: Whom shall the smart home serve?

Footnote May 2021

Some legal advice for landlords just in from US law firm Orrick: "Tenant data may be an attractive source of new revenue, but landlords should proceed with caution" (13 May 2021). They also note that "New York City Council has enacted a Tenant Data Privacy Act that is poised to enhance privacy protections in multifamily buildings in the city" (27 May 2021).

Dieter Bohn, Situation: there are too many competing smart home standards. Surely a new one will fix it, right? (The Verge, 19 Dec 2019)

Ross Clark, The critics of smart meters were right all along (Telegraph, 19 September 2020) HT @tprstly

Constellation, Smart Homes Allow the Aging to Maintain Independence (published 20 July 2018 updated 13 August 2018)

Kayla Devon, The Lure of the Smart Apartment (MFE, 31 March 2016)

Karen Jordan, Set It And Forget It: The Lure Of Smart Apartments (Forbes, 28 August 2017)

Kieren McCarthy, The IoT wars are over, maybe? Amazon, Apple, Google give up on smart-home domination dreams, agree to develop common standards (The Register, 18 Dec 2019)

Michael McKee, Your Landlord Could Know That You’re Not at Home Right Now (New York Times, 17 December 2019)

Alfred Ng, Smart home tech can help evict renters, surveillance company tells landlords (CNET, 25 October 2019)

Ashlee Clark Thompson, Persuading your older parents to take the smart home leap (CNET, 11 April 2017)

Shannon Yavorsky and David Curtis, Unlocking the Value of Tenant Data (Orrick 13 May 2021), Home Alone? New York City Enacts Tenant Data Privacy Act ( Orrick 27 May 2021) HT @christinayiotis

Related posts: Understanding the Value Chain of the Internet of Things (June 2015), Defeating the Device Paradigm (Oct 2015), Hidden Functionality (February 2019), Towards Chatbot Ethics - Whom does the chatbot serve? (May 2019), Driverless cars - Whom does the technology serve? (May 2019), The Road Less Travelled - Whom does the algorithm serve? (June 2019)


 Updated 16 November 2020, 29 May 2021

Friday, October 11, 2019

Insights and Challenges from Mulesoft Connect 2019

#CONNECT19 @MuleSoft is a significant player in the Integration Platform as a Service (iPaaS) market. I've just spent some time at their annual customer event in London.

Over the years, I've been to many events like this. A technology company organizes an event, possibly repeated at several locations around the globe, attended by customers and prospects, business partners, employees and others. After an introduction of loud music and lights shining in your face, the CEO or CTO or honoured guest bounces onto the stage and provides a keynote speech. Outside, there will be exhibition stands with product demonstrations, as well as information about complementary products and services.

At such events, we are presented with an array of messages from the company and its business partners, with endorsements and some useful insights from a handful of customers. So how to analyse and evaluate these messages?

Firstly, what's new. In the case of Mulesoft, the core technology vision of microserves, networked applications and B2B ecosystems has been around for many years. (At the CBDi Forum, we were talking about some of this stuff over ten years ago.) But it's useful to see how far the industry has got towards this vision, and how much further there is to go. In his presentation, Mulesoft CTO Uri Sarid described a complex ecosystem that might exist by around 2025, including demand-side orchestration of services. There is a fair amount of technology for supply-side orchestration of APIs, but demand-side orchestration isn't really there yet.

Furthermore, organizations are often cautious about releasing the APIs into the wild. For example, government departments may make APIs available to other government departments, local governments and the NHS, but in many cases it is not yet possible for citizens to consume these APIs directly, or for a third party (such as a charity) to act as a mediator. However, some sectors have made progress in this direction, thanks to initiatives such as Open Banking.

However, the technology has allowed all sorts of things to be done much faster and more reliably, so I heard some good stories about the speed with which new functionality can be rolled out across multiple endpoints. As some of the technical obstacles are removed, IT people should be able to shift their attention to business transformation, or even imagination and innovation.

MuleSoft argues that the API economy will drive / is driving a rapid cycle of (incremental) innovation, accelerating the pace of change in some ecosystems. MuleSoft is enthusiastic about  citizen integration or democratization, shifting the initiative from the Town Planners to the Settlers and Pioneers. However, if APIs are to serve as reusable building blocks, they need to be built to last. (There is an important connection between ideas of trimodal development and ideas of pace layering, which needs to be teased out further.)

Secondly, what's different. Not just differences between the past and the present, but differences between Mulesoft and other vendors with comparable offerings. At a given point in time, each competing product will provide slightly different sets of features at different price points, but feature comparisons can get outdated very quickly. And if you are acquiring this kind of technology, you would ideally like to know the total cost of ownership, and the productivity you are likely to get. It takes time and money to research such questions properly, and I'm not surprised that so many people rely on versions of the Magic Sorting Hat produced by the large analyst firms.

By the way, this is not just about comparing Mulesoft with other iPaaS vendors, but comparing iPaaS with other technologies, such as Robotic Process Automation (RPA).

And thirdly, what's missing. Although I heard business strategy for APIs mentioned several times, I didn't hear much about how this could be done. Several speakers warned against using the term API for a non-technical audience, and advised people to talk about service benefit.

But how to identify and analyse service benefit? How do you identify the service value that can be delivered through APIs, how do you determine the right level of granularity and asset-specificity, and what are the design principles? In other words, how do you get the business requirements that drive the use of MuleSoft or other iPaaS products? I button-holed a few consultants from the large consultancies, but the answers were mostly disappointing.

I plan to attend some similar events this month, and shall write a couple of general follow-up posts.


Here's an article I wrote in 2002, which mentioned Sun Microsystems' distinction between micro services and macro services.

Richard Veryard, Identifying Web Services (CBDI Journal, February 2002)

And here are two articles discussing demand-side orchestration.

Richard Veryard and Philip Boxer, Metropolis and SOA Governance (Microsoft Architecture Journal, July 2005)

Philip Boxer and Richard Veryard, Taking Governance to the Edge (Microsoft Architecture Journal, August 2006)

See also CBDI Journal Archive

Link to MuleSoft presentations

Related post: Strategy and Requirements for the API Ecosystem (October 2019)