We have just crossed a threshold – it’s time to fix Facebook, Twitter et al.

Myanmar genocide

We have crossed a threshold where technology Facebook et al are potentially facilitating fixed elections and genocide. It’s time to act.
At long last there is wider interest in the Facebook and Twitter models and growing understanding of the damage they are doing to society. We have all seen the widespread interference in elections by state actors and vested interests; the rise of angry politics in the USA, UK and elsewhere, channelling discontent and splitting society; the spread of fake news; the reduction in trust generally and support for criminal or terrorist based activities; and so, the list keeps getting longer. Last week however I felt a threshold was crossed that needed to be exposed.
Last Tuesday Kevin Roose wrote in the New York Times about the activities of Facebook in Myanmar and other parts of the developing world. He explains in 2016 Facebook joined a partnership with MTP, the state-run telecom company to give subscribers access to its Free Basics program. Free Basics includes a suite of internet services including Facebook, that can be used without counting towards the cellphone data plan. As a result, the number of Facebook users in Myanmar has skyrocketed from 2 million in 2014 to more than 30 million today. Apparently violence against the Rohingya Muslims has been fuelled in part by misinformation spread by Facebook which is now used as a primary news source by many people in the country. Doctored photos and unfounded rumours have gone viral on Facebook, many of them shared by official government and military accounts. I’m sure there’s more to the ethnic cleansing happening as we speak in Myanmar, but at the very least Facebook can clearly be accused of facilitating genocide!
As Facebook, Twitter and Google were grilled last week by the USA Congress, all the attention was on interference in presidential elections by Russian state actors. Whilst this is fully understandable, I suggest it’s insufficient. The intelligence emerging from Myanmar tells us the platform model is a clear threat to society and urgent response and actions are necessary. Even though the platform companies are US corporations, given the state of politics in the USA, I suggest we shouldn’t wait for the USA to fix the problem. And it goes without saying that the platform providers, Facebook, Twitter and Google are completely incapable of exerting the governance because their business models are in direct conflict with appropriate actions.
I blogged in August under the title, It’s time to exert governance over the Global Tech Leaders!
But even though my thinking was probably radical for many, it didn’t go far enough. I believe now there is just one primary focus required – to make the platforms trusted. Some key actions include:
1. All three platforms must provide a Trust Review button where all posts, tweets and search responses can be rated for reliability on several dimensions, including Ownership and Validity. The objective is to ensure transparency of authorship and elimination of anonymity, and trustworthiness of content and elimination of fake news.
2. All users are encouraged to provide review feedback. Think Trip Advisor, where reviewers themselves are reviewed and rated.
3. All platform providers are required to provide open (service API based) access to content to allow external and independent actors to develop real-time AI based review engines, that contribute to the trust rating.
We might imagine that the platform providers themselves would be motivated to provide AI based Trust engines, but so would reputable communications companies, perhaps in conjunction with platform providers.

How might this happen? First it seems to me that the platform providers might just realize this is necessary and act because these actions would actually help their business model, and even provide empirical data that allows them to compete more effectively. Second, public opinion; we need to energize platform users to help them understand the damage that’s being done to society. In my earlier post on this topic I considered the UN as a point for governance; and I still believe this is the best long term solution. But right now we need action fast and triggering better behaviors from the platforms themselves is probably the best course of action that will protect their business model as well as society as a whole.

Advertisements
Posted in Artificial Intelligence, Facebook, Twitter, Google, Governance, Technology Platforms, Trust | Leave a comment

Was Hurricane Ophelia Damage Avoidable?

Ophelia 2

This week Ex-Hurricane Ophelia directly struck Ireland causing widespread damage and chaos. This is regarded as an exceptional event, the worst storm in recorded Irish history. The National Emergency Coordination Group  (NECG) were reported in the Irish Examiner as saying, “Trees falling and taking down powerlines is the main cause of power outages and road blockages. At the peak 385,000 homes and businesses were without power.” The chaos and disruption has been extensive, and the ESB (Electricity Supply Board) advise it will take up to 10 days to reconnect all customers.

So, is this an act of God, or should we have been better prepared? Of course there were tragic outcomes and much damage that was clearly unavoidable. But we need to examine whether the widespread disruption to peoples’ lives was actually an acceptable outcome.

I live in West Cork – directly facing the Atlantic and aptly named Roaring Water Bay; one of the areas hardest hit by Ophelia. Back in the early noughties I had a visit from an ESB maintenance crew. They arrived without notice and, while being very polite made it perfectly clear they had the right to enter my property and lop the top off a tree that in their opinion was overhanging the electricity supply. I had no issue with this and the problem was quickly sorted and they went on their way.

Later on I learnt that the ESB employ linesmen, engineers that literally walk the overhead supply network checking the integrity of the system, identifying potential problem tree growth or loose poles. Remember the song, The Witchita Linesman? You get the idea.

Now fast forward to today. After Hurricane Ophelia had passed over I examined that same tree that the ESB crew trimmed some 15 years ago. Sure enough it appeared to me that once again it would represent a real threat to the supply line if it came down. So I ask myself, “where are the ESB linesmen?”. Given the uncontroversial statement that “trees falling and taking down the powerlines” is the main cause leading to days or weeks inconvenience or worse, why were these trees not subject to normal ESB linesman checks and maintenance action?

We need to ask the ESB whether the spend on linesmen and maintenance crews has remained static or increased or decreased over the past couple of decades. Or whether they made explicit policy decisions to reduce planned maintenance to improve the ESB accounts on the basis that one time exceptional cost might be treated differently, perhaps by government intervention or insurance.

Hurricane Ophelia is clearly a natural phenomenon. But the impact and amount of disruption are not. I heard the responsible government minister speaking on radio blaming both cause and effect on Climate Change, suggesting we need a huge infrastructure budget over the next decade or so to ensure this doesn’t happen again. I suggest he is entirely incorrect. While we can never avoid tragic accidents and incidents of major damage, we can mitigate the worst effects of the storm by proper maintenance. The question is, have the ESB actually made matters worse by reducing linesmen and maintenance crews to reduce costs and resources. If so they should be held to account.

Posted in Uncategorized | Leave a comment

Do we really need MORE democracy? Is consent of the people a recipe for instability?

Farage

I was intrigued to see today that as Jean-Claude Junker addressed the European Parliament he was heckled by Nigel Farage who said, “what was being proposed was more Europe in every single direction… without the consent of the people”. It prompted me to consider how democracy is failing in the UK and the USA, precisely because the technology enabled voice of the people has driven populist agendas which have produced highly volatile outcomes and increasingly uncertain futures. Mr. Farage wouldn’t like it said, but because is the EU is all about rules and processes it can provide a strong layer of stability and predictability, which frequently exerts governance over the more volatile actions of national governments. While UK Brexiteers may be upset by the difficult negotiation process currently in process, it is the EU team that is forcing the British to address the intricacies and practicalities involved in extricating themselves from Europe. Note this governance is not only happening in the UK, but also Romania and Poland that are pushing the democratic envelope as their right-wing governments test their authority.
The evident failure of democracy in the UK and USA, captured by populist agendas and misled by fake news suggests we need a counter-balance. Certainly, in this global era we badly need supra-national governance to manage global policy issues such as macro finance, trade, security, data etc. One might even be forgiven for thinking that some form of benevolent autocracy might be an improvement over the likes of Donald Trump, Theresa May, Nigel Farage and Boris Johnson etc. But wait a minute, that’s exactly what we have in the EU. It’s not perfect, we know that. The EU actually represents a supra-national governance body that has demonstrably delivered stability and peace, well apart from the banking crash, and represents a blend of technocratic and democratic government which I think is likely to be preferable to say, the Chinese system.
So why reinvent a wheel that is actually working? The EU is needed. If it wasn’t in place, we would need to invent it. What we do need to do is make it work better.

Posted in Democracy, Governance, Uncategorized | Leave a comment

It’s time to exert governance over the the Global Tech Leaders!

Tech Governance 2

Nearly 20 years ago, in my research work I and colleagues identified the IT platform companies as implementing a radical new business model that would be profoundly transformational in many dimensions, in terms of potential for astonishing levels of growth, global market share, monopoly and competitive lockout and opportunities for small companies to slingshot their capability onto the global stage.
These things have come to pass. However, what we didn’t foresee was the potential for abuse – specifically the collection and resale of individuals’ data, the profiling and tracking for commercial advantage and more recently for political advantage. Also, the use of the platform for nefarious purposes. Nor did we identify the issue of split responsibility and governance in a platform environment.
All enterprises require some form of governance, internal and external. In the past external governance of corporations has been about ownership, assets and financial probity. More recently there has been some moves to implement more qualitative forms of governance to require corporations to report on issues such as climate change and social responsibility. Clearly, we now need strong governance over the technology companies and their customers use of the platform. I am speaking specifically about Facebook, Twitter and Google as priorities, but I fully anticipate the requirement will be much broader.
Let’s consider a simple model.
1. Establish a global governance board under the aegis of the United Nations. Allow the UN to appoint acknowledged tech sector experts who have unquestioned neutrality to a governance board with a limited term of 3 years. Charter and resource the board to propose and publish policies and monitor and report 6 monthly on key compliance issues.
2. Example policies:
a. Establish limits over individuals’ or enterprises’ use of the social platforms. Require named tech companies to make an account charge of $1 per month for every follower/friend over say 1000.
b. Require platforms to highlight paid adverts and accounts to other users.
c. Require named platforms to provide open standards for interaction between social platforms to reduce monopoly behaviors.
d. Require user ownership of personal data and rights to opt in or out of profiling based messaging and advertising.
e. Require message privacy to be subject to external governance by approved security services.
f. Recommend country level taxation systems that claw back excess profits supplementary to corporate tax regimes.
3. Charter the UN governance board to establish communications with country level governance boards and facilitate coherent implementation of policies at global and country level.
These are simply ideas. The biggest issue will be, how to make it work? Because of course there are no global governance bodies that would see this as their responsibility. Further national governments are restricted by various treaties on cross country taxation. However, we do have examples of global governance. It must be said that climate change is perhaps not a good example to copy. A better example is the Montreal Protocol on Substances that Deplete the Ozone Layer. This is an international treaty designed to protect the ozone layer by phasing out the production of numerous substances that are responsible for ozone depletion. It was agreed on 16 September 1987, and entered into force on 1 January 1989, followed by a first meeting in Helsinki, May 1989. Due to its widespread adoption and implementation it has been hailed as an example of exceptional international co-operation, with Kofi Annan quoted as saying that “perhaps the single most successful international agreement to date has been the Montreal Protocol”.
Further, we might expect that the tech companies in question may be prepared to cooperate in some form of self-governance on key policies that could be implemented very rapidly. Certainly some of the tech leaders have emphasized the importance of moral compass. Google has recently moved from “don’t be evil” to “do the right thing”, and Facebook’s Mark Zuckerberg has made various socially responsible comments including, “give people power”. If the tech companies are smart, and they usually are, they will see they are better off negotiating rather than having governance forced on them.
Very clearly the time has come that we need to be sure political processes cannot be abused, that individuals’ privacy is respected and that profits made by tech companies are taxed appropriately. If demagogues or manipulators can afford to “pay” for the privilege of having 35 million followers, then at least there will be visibility of the cost benefit. These are problems that can and should be fixed, now.

Secrets of Silicon Valley, BBC2

Montreal Protocol

Technology and Society – Problem or Opportunity

 

 

Posted in Digital Transformation, Governance, Uncategorized | 1 Comment

Terrorism requires new thinking about Technology in Society

Surveillance Image

The abominable attacks in the UK, France and Germany are clearly unacceptable; but the responses of both security services and political leaders are completely inadequate to these types of threat.
Today we hear that one of the London Bridge attackers was well known to the police and media. But “there was insufficient evidence he was part of a plot, so was not actively investigated”. Also, one of the same group of attackers had been hiding in plain sight in Dublin, travelling freely between the UK and the Republic, presumably associating with other similarly radicalised individuals. Perpetrators of the Manchester attack had evidently been identified as radicals by members of the public who informed the police, who evidently didn’t or were unable to respond. And the UK police at last admit they have insufficient resources.
We must sympathize with the police and security services; they have an impossible job; the terrorists can live freely in society and increasingly use everyday artifacts as weapons such as rental vans and trucks and knives. Conventional policing where suspects are tracked involving huge resources is never going to have enough resources.
A conventional response, which was used in times past with the IRA, would be to inter individuals that were clear threats where there was no hard evidence. This was always highly controversial and goes against all our democratic norms. And of course, we could perhaps inter individuals returning from Syria or Iraq who were known to have been involved in those conflicts. But for how long? And what about those that didn’t travel but provided resources, funding and other forms of help?
A modern approach would be to track patterns of behaviour. We are all aware that everything we do is tracked, including telephone calls, emails, social media posts and messages as well as purchases, travel arrangements, border crossings etc. as well as particularly in the UK, CCTV surveillance. Terrorists will have particular behavioural patterns – associations with others who are also suspected, travel to particular locations, use specific web sites, attend specific churches etc. This technology is available and mature; in active use by commercial organizations and tech companies involving sophisticated use of big data and artificial intelligence. But we must assume that the police and security services are not using these systems, at least in any effective manner. Perhaps they are constrained by law?
Strangely I don’t hear or see comment on this aspect of how we should respond. Why? Are we too precious about our democracy? The fact is our democracy is already being undermined by political operators and indeed politicians that have swung and radicalized opinion because of threats to society. It’s time we grasped this challenge and accepted some loss of privacy in order to protect society. It’s way past the time for political leaders to use vague statements of intent; or to focus resourcing as the only solution. We should be using the power of the very latest technology and requiring the tech companies get deeply involved in addressing all the issues that have been talked about for too long and not acted upon. And implement across Europe.
It’s evident the police and security services are simply plodding. They need to be brought into the 21st Century.

Posted in Artificial Intelligence, Uncategorized | Tagged , | 1 Comment

Technology and Society – Problem or Opportunity

In economics Kondratiev waves (or supercycles) are cycle-like phenomena in the modern world. [1] Initially identified in 1925 as major economic cycles by Nikolai Kondratiev, Joseph Schumpeter suggested in 1939 naming the cycles “Kondratieff waves” in his honor. The theory originally hypothesized the existence of very long-run macroeconomic cycles lasting between 50–54 years. Kondratiev identified three phases in each cycle: expansion, stagnation, and recession. More common today is the division into four phases with a turning point (collapse) between the first and second phases. This latter phase seems very realistic, and is well articulated in the work of Geoffrey Moore in Crossing the Chasm [2].
According to the innovation theory, these waves arise from the clustering of foundational innovations that trigger technological revolutions that in turn create new industries and or commercial markets. However, it is also important to understand that there are also close relationships between technology and economic cycles which may accelerate or reduce adoption and therefore cycle length.
There are several modern versions of the technological cycles. They may be generalized as follows:
• Industrial Revolution—1771
• Steam and Railways—1829
• Steel and Heavy Engineering—1875
• Oil, Electricity, the Automobile and Mass Production—1908
• Information and Telecommunications—1971
And these hypothesised cycles seem to support the cycle duration theory in the range 50 – 60 years. And while we are forever hearing that change gets more and more rapid, we should remind ourselves that the four phases identified by Kondratiev and Schumpeter exert natural constraints on the phase and cycle duration timescale – formation and evolution of foundational R&D, infrastructure, enterprises, markets, and supply chains takes time.
This basic technology cycle model may help us to understand what’s happening in our world today. Some key questions. Are we right now in an Information and Telecommunications cycle? What phase? Is the Internet a phase of this, or is there an Internet cycle? I am minded to say that this phase is reasonably named, and it is in stagnation – that our ability to leverage the foundational information and telecommunications infrastructure is over stretched. And that a new cycle of Artificial Intelligence is in the very early stages of the expansion phase which leverages the essential information and telecoms infrastructure to enable a huge array of application areas in all vertical sectors based on reasoning, decisioning, differentiated user experience, robotics etc.
Let’s examine a little more closely the stresses in information and telecommunications that indicate we are in the recessionary phase. First we can see that technology is having major impacts on society. While social media has an extraordinary and pervasive level of adoption [3] the capabilities are being widely abused. As a result social media is directly impacting on society, facilitating corruption of electoral processes, information integrity and opinion, often creating popular anger and influencing various outcomes in elections, referenda, opinion polls etc. We will all be aware how Donald Trump very effectively uses Twitter to apply pressure in a very public manner. You might conclude that the concepts of big data are being used for unintended purposes.
In response to these obvious flaws the tech companies have responded. But the responses are very tactical, to identify and remove fake news, bullying and criminal behaviors and stop abuse. This is purely tactical; the absence of standards and legislative governance is clearly a major problem, exacerbated by the global nature of the tech industry at a when many countries are rejecting globalism for nationalism, very possibly as a direct result of the abuses. And this chaotic behaviour is only the start. As we move deeper into the Artificial Intelligence technology cycle, the pressures on society will become extreme. Already many are concerned that AI will eliminate many jobs, and not just the mundane and routine, but many skills and disciplines that can be hugely improved using vast stores of knowledge of best practice.
It seems the tech companies are attempting to respond to some of the most pressing symptoms of anarchic data management. But, notwithstanding brave mission statements such as “do no harm”, the tech companies cannot reinvent society themselves. What’s required is a new model for society that provides a backplane for tech innovation. Instead of tech companies innovating simply to make more money in the next 4 quarters, what’s needed is a vision of how tech can support dramatic changes in the way we live, work and play. And in the longer term this approach is likely to deliver a more stable transformation of society into the AI based technology world which attempts to match educational systems with likely resource requirements in the appropriate timeframe. Standards that facilitate integrity, security and privacy, protecting the individual and society.
Of course, there’s a major problem that the greater the level of AI based automation, the greater the probability that we will deliver a big brother like society. And we can observe right now a great fragmentation of society as technology gives more and more people a voice, yet conventional governance systems prevent wide scale involvement apart from voting for individuals or parties every few years. No surprise people feel they are becoming more and more disenfranchised and therefore more marginalized, and are we surprised when radical politicians such as Trump, or radical ideas such as Brexit gain traction; mostly it must be said in protest against the status quo.
So part of the vision and new model needs to facilitate new role and responsibility systems. Many of us will be very familiar with Agile projects. The core idea that small teams, given responsibility and accountability are the ideal way to deliver something useful. Perhaps this is a model that we should be looking to use more widely.. Some examples of decentralization and distribution:
– Back in the 1990s in the UK, many schools were released from local government control; the “grant maintained” schools ran themselves. Albeit with exceptionally strong central governance in the Ofsted system. As the time I will admit I thought it highly improbable that parent based governing committees could rise to the task; but I was proved wrong – in practice and now for decades these independent schools have delivered excellent education with very high levels of local management.
– Recently I noted, again in the UK, a Labour politician was elected as mayor of the Greater Manchester. Whilst Andy was and is a very senior Labour party politician, he rejected most of the central party policies, and ran his campaign specifically on key local issues. In effect he was saying, Greater Manchester is different; we need localized policies that will work for us. He was elected in a landslide.
Consider now, what could we do with tech to facilitate all manner of distributed and decentralized enterprises and projects? Frankly the social media tools today are focused on making money for the tech company and its platform participants. Plainly they do not support collaboration, education, team building and management whether formally or informally. Sure, they can be useful at some level; but consider what could be done if enabling and facilitating meaningful dialogue, higher integrity of all types of data, appropriate levels of AI based governance, coordination of distributed activities encouraging local or distributed groups to form for defined purposes, was at the heart of the platforms? What’s the business model? If governments saw that well organized local groups were a force for good, overcoming some of the most difficult problems in society through concerted action, we might imagine that government funding would be forthcoming. Perhaps governments could even work with tech companies on supporting localized solutions for big national problems that really only have localized solutions.
I accept this is going against the grain of globalization; but globalization and localization should work side by side; small groups collaborating with other small groups to share ideas, practices work products etc. as well as forming standards bodies that provide guidance for wider communities.
The tech companies are at the forefront of societal change. My thesis is that they need to take some responsibility for the outcomes. Many of the tech leaders have made fortunes from their efforts; some have become philanthropists, usually in very different sectors from tech. I suggest it’s time for some of these folks to address their philanthropy to solve the problem that they are handing to society; otherwise we will have a dystopian future of their own making!

[1] Kondratiev wave

[2] Crossing the Chasm

[3] Number of social media users worldwide from 2010 to 2020 (in billions)

Posted in Artificial Intelligence, Digital Transformation, Uncategorized | Tagged | 1 Comment

Optimizing User Experience AND Customer Satisfaction with AI?

Let me tell you a true story. A couple of weeks ago I received an email reminding me to renew my motor (auto) insurance. I wasn’t best pleased as the premium had increased considerably. So I called the insurance company. Speaking to the agent he offered me 5% discount, as if it was a personal favour he was doing for me. I said, “that’s exactly what happened last year. Why don’t you just build that discount into the rate for long term customers?” And I continued, “But more importantly, even with the 5% off the top, the premium has now increased by over 20% in two years!” The agent said basically, “take it or leave it.” I asked for the case to be escalated on the basis of being a long standing customer, but he became just a little stressed. So I wrote an email to the firm, stating my case; and after two days I received a standard reply saying, sorry nothing doing. You might be thinking, why didn’t you look around – and I did, but I have several policies with this firm and I want to keep the multi-policy discounts.

At this stage the renewal is coming within 7 days, and the very next day I receive an email making me some amazing offers if I renewed; would you believe a €200 holiday voucher, or a three month cinema pass, or fully funded family fun day, or a round of gold on a leading course. Now none of these would really cost the insurance company €200, but they would still involve serious cost, even if the uptake rate post renewal was low.

So I called the company again and spoke to another agent, “why not just cut the cost of my premium by say half of what the special offer would cost?” Predictably the answer was in the negative. Oh just not possible. Nor was it possible to talk to someone who might just be interested in what a customer had to say!

I make no apologies for the shaggy dog story because it highlights an essential truth about how large businesses operate. They fail to coordinate marketing and operations in both cost and policy. So when business analysts talk about cross channel user experience are they simply figuring out the mechanics? Do they ignore the policy separation and customer behavioral impact? Clearly that’s what happens in this big insurance company, and I would bet they are typical. They take a view that they will lose a certain percentage of their base on renewal each year and work on new business strategies to compensate. As opposed to looking at the overall problem from the customer behavioral perspective!

What’s this got to do with AI? Well you might say that your AI based system is only as good as your policy set and your ability to drive policies and strategy from an understanding of customer behavior. We have all complained down the years about the literally terrible call center agent experience; but very often this is not down to skills, it’s down to policy implementation, not allowing the agent to operate as a human by taking responsibility for some discretion. So in an AI context the question is with what sort of intelligence will you aim to replace your agents? And how much do you really understand about your customer behaviors so you can implement customer responsive policies?

To conclude, user experience (UX) must be more than just the mechanics of allowing the customer to renew a policy, place  an order, manage a process etc. It’s got to respond to real customer needs in a manner that improves the business performance and customer satisfaction. And this won’t happen without major change in the way you track customer needs and develop policy. And surprise, surprise, this will lead to AI getting a bad name, not just for destroying jobs but also being totally unresponsive as more and more barriers are put in place between the organization and it’s customers.

Posted in Artificial Intelligence, User Experience | Leave a comment