Do You Know You Are Using And/Or Exposed To Artificial Intelligence?

Source: datamation

Amid the news that the Department of Work and Pensions is using artificial intelligence to decide benefits claims, a third of UK financial firms recently told the Bank of England and Financial Conduct Authority they were not using 'machine learning' (a subset of AI) in their businesses. That's pretty concerning, given the kind of AI we all use without realising and the fact that anyone wrongly denied benefits will need short term finance. That response also begs the question whether those firms know how their many stakeholders are using AI (whether unwittingly or not). If their suppliers, intermediaries, customers, investors, competitors and the government are using AI, then how does that affect their own strengths, weaknesses, opportunities and threats? And how does that in turn affect their stakeholders? No city on Earth is ready for the disruptive effects of artificial intelligence. Also worrying is the finding that smaller, fintech firms seem to believe that machine learning is no use to them. And given the problems with AI explained below, it's important for everyone to consider whether and how they rely on or are otherwise exposed to the two thirds of financial firms who are actually using AI... hyping the benefits without understanding the shortcomings will harm our trust AI where it could be helpful.

What is AI?

The term "AI" embraces a collection of technologies and applications, with machine learning usually featuring at some point:

• narrow artificial intelligence
• machine learning
• artificial neural networks 
• deep learning networks 
• automation
• robotics
• autonomous vehicles, aircraft and vessels
• image and facial recognition
• speech and acoustic recognition
• personalisation
• Big Data analytics

At the recent SCL event I chaired in Dublin, Professor Barry O'Sullivan explained that AI technologies themselves may be complex, but the concepts are simple. In essence, machine learning differs from traditional computer programming in that:

• traditionally, we load a software application and data into a computer, and run the data through the application to produce a result (e.g. how much I spent on coffee last year);

• while machine learning involves feeding the data and desired outputs into one or more computers or computing networks that are designed to write the programme (e.g. you feed in data on crimes/criminals and the output of whether those people re-offended, with the object of producing a programme that will predict whether a given person will re-offend). In this sense, data is used to ‘train’ the computer to write and adapt the programme, which constitutes the "artificial intelligence".

What is AI used for?

AI is used for:

1. Clustering: putting items of data into new groups (discovering patterns);
2. Classifying: putting a new observation into a pre-defined category based on a set of 'training data'; or
3. Predicting: assessing relationships among many factors to assess risk or potential relating to particular conditions (e.g. creditworthiness).

The Challenges with AI

The primary concerns about AI relate to:

1. cost/benefit - $50m in electricity to teach an AI to beat a human being at Go, hundreds of attempts to get a robot to do a backflip, but it can't open a door;
2. dependence on data quantity, quality, timeliness and availability;
3. lack of  understanding - AI is better at some tasks than humans (narrow AI) but general AI (same as humans) and superintelligence (better than humans at everything) are the stuff of science fiction. The AI that can predict 79% of European Court judgments doesn't know any law, it just counts how often words appear alone, in pairs or fours;
4. inaccuracy - no AI is 100% accurate;
5. lack of explainability - machine learning involves the computer adapting the programme in response to data, and it might react differently to the same data added later, based on what it has 'learned' in the meantime; 
6. the inability to remove both selection bias and prediction bias - adding a calibration layer to adjust the mean prediction only fixes the symptoms, not the cause, and makes the system dependent on both the prediction bias and calibration layer remaining up to date/aligned over time; 
7. the challenges associated with the reliability of evidence and how to resolve disputes arising from its use; and
8. there's a long list of legal issues, but lawyers aren't typically engaged in development and deployment.

This means the use of AI cannot be ignored. We have to be careful to understand the shortcomings and avoid hyping the benefits if we are to ensure trust in AI. That means challenging its use where the consequences of selection bias or false positives/negatives are fatal or otherwise unacceptable, such as denying fundamental rights or compensation for loss.

Instead, artificial neural networks and deep learning are better used to automate decision-making where "the level of accuracy only needs to be "tolerable" for commercial parties [who are] interested only in the financial consequences... than for individuals concerned with issues touching on fundamental rights."

Being realistic about AI and its shortcomings also has implications for how it is regulated. Rather than risk an effective ban on AI by regulating it according to the hype, regulation should instead focus on certifying AI's development and transparency in ways that enable us to understand its shortcomings to aid in our decision about where it can be developed and deployed appropriately.

The Conspiracy Theory Of Conspiracy Theories

I'm fascinated by conspiracy theories, and just happened across the Daddy. Corey Doctorow, the entertaining science fiction writer, believes he's found the source of them all:

"... [conspiracy theories] aren’t attributable to ideology or mind-control rays... at root, they are a disagreement about how we know whether something is true or not. When we argue about the flat Earth, we’re not just debating the shape of our planet: We’re debating the method by which we can know its shape....40 years of rising inequality and industry consolidation have turned our truth-seeking exercises into auctions, in which lawmakers, regulators and administrators are beholden to a small cohort of increasingly wealthy people who hold their financial and career futures in their hands. Industry consolidation makes it startlingly cheap to buy the truth: Once an industry consists of a handful of players, it’s easy for everyone to agree on the play, and the only people qualified to be referees are drawn from the companies’ own executive ranks, whence they will return after their spin in governance’s revolving door."

Not only is this thesis merely political hogwash designed to justify a call to break up certain industry oligopolies, but it also serves to highlight how the human brain works to make conspiracies attractive in the first place (as explored by Nobel prize-winner Daniel Kahneman and Nicholas Taleb, often mentioned here). 

The central problem is 'overconfidence' - the belief that our existence is somehow 'controlled' rather than random (also at the root of creationism vs natural selection, for example).

I mean, just imagine if all the data that every human perceives really were "bought" or curated by "a handful of players" - let alone after only 40 years of trying...

A moment's thought, let alone a cursory read of James Gleick's excellent history of humanity's attempt to master information reveals this would not be possible. And many different conspiracy theories abound in the many different communities around the planet.

Even complaints about deliberate misinformation causing the outcome of specific votes, like Brexit, are overdone. As Taleb points out, such outcomes may be classified as Black Swans - surprise events that we try to rationalise by hindsight. Lies and misinformation might be part of the swirl surrounding Brexit, yet may really only correlate with the result without being causative - they could be just as symptomatic of a deeper malaise as the result of the vote itself. 

That's not to say lies and electoral violations should be overlooked or go unpunished - or that we should not try to rectify obvious economic mistakes or life-threatening misconceptions like the "Anti-Vax" movement. But ultimate 'control' just isn't an option. We can only minimise our exposure to the downside, and maximise our exposure to the upside, of the major events that shape our history.

Applying that to Doctorow's theory: the break of up of industry oligopolies may be worthwhile for certain reasons, but that process will not somehow restore human control over the 'truth' - any more than it will alter human psychology.

Why Suggest A Bridge Between Scotland and Ireland?

Popularity develops through the snowball effect and the bandwagon effect. Some things or people are popular because they're useful, solve a real problem or are widely appreciated for some intrinsic quality: the wheel, a rock song/group, the telephone...  But what if you have nothing genuine to offer? How do you create your own snowball and bandwagon effects out of thin air? 

You do what every snake oil salesman in the Wild West did, what Bernie Madoff did, what every religious sect leader has done, what reality TV show producers do - and what the likes of Farage, Johnson and Gove have done...

You create or 'boost' something hugely ambitious, unique, deceptively simple, vaguely plausible but not actually 'real' in the sense of being achievable or demonstrable. Something in which your victims can only have faith.

Committing to solve a really big, actual problem is out of the question. Firstly, it's really hard and will take a long time, because it involves changing actual behaviour and fighting inertia - the resistance to change. Secondly, it won't make you exclusively popular because you'll likely be competing with lots of others trying to solve the same problem. You won't stand out in the crowd. You won't capture peoples' attention.

And real problems won't capture your followers' imagination or their faith.

To make your victims cling to their belief for a long time, you also need a goal that’s really big and ambitious - because, if they go for it, a big project will commits lots of people, money and resources that can't easily be redeployed.

And if the goal isn't 'real' it can never quite be achieved or delivered or found (think the Holy Grail): the quest is never-ending. Once you’ve got your popularity snowballing it's just about the journey, not the destination. Nobody wants to stop believing (and spending). Negotiating trade deals is an endless process - the job of leaving the EU successfully will literally never be "done". There's no easy way to leave the cult. Inertia is now on your side!

Here are some examples of projects that were or are hugely ambitious, deceptively simple, vaguely plausible but not actually achieved or achievable that seem to have been conjured up to boost someone's popularity... You'll see that Johnson's proposal for a bridge between Scotland and Ireland fits right in - get that feasibility study started!

• an airport in the middle of the Thames;

• long term stock market returns of 11% (not the widely accepted 8%);

•  Farage's decades long quest for Britain to leave the EU;

• the 'Leave' campaigns, as backed by Johnson/Gove and Farage/Banks;

• the garden bridge;

• donate X% of your income for eternal salvation;

• build a bridge between Scotland and Ireland. 

And finally, a word from the perpetrator of the biggest fraud of all time... (or is it?)...

Which UK Government Policies Would Prevent Free Trade Deals?

We know that the Johnson regime wants a No Deal Brexit, and that Labour only opposes that subject to agreeing its own form of Brexit. But any form of Brexit would doom the UK to lengthy trade talks and related political turmoil in a situation where Brits can no longer live and work freely in 30 other countries or trade freely with the EU and other countries under EU free trade deals. A limited 'transition' or 'implementation' period merely sugar coats the pill: Brexit spells the end of freedom of movement for labour, goods, services and capital.

In this context, the prospects for new free trade arrangements are absolutely critical, and a key consideration is whether any UK government policies or party conference 'manifesto commitments' stand in the way of negotiations. 

I've started a general list below, including policies that could either directly prevent a deal being agreed and those that could discourage businesses from a partner country actually investing in the UK or undertaking trade anyway:

• breaching the Good Friday Agreement - the US congress has already signaled that it will not approve a free trade deal with the UK that endangers the GFA;

• nationalisation plans - EU state aid rules already permit nationally owned assets (though rules on nationalising/subsidising railways are changing, for example) but Labour has said that it wants to nationalise a much wider scope of the UK economy and "rejects the whole principle of free trade";

• capital controls: the UK government might first introduce controls aimed at preventing the flight of funds resulting from policies that are not compatible with existing foreign investment or that will make the UK uncompetitive with other investment opportunities. Nationalisation plans would be a particular problem given the amount of foreign ownership of utilities and other formerly public assets. There must also be concerns that foreign nationals will want to get their assets out when they wish to leave (and before any capital controls come in). Capital controls are also part of a country's monetary armoury for numerous other reasons - especially when an economy is teetering on the brink of disaster. Capital controls are excluded in US free trade deals and would likely be contested in international arbitration, subject to the IMF's power to require them to safeguard its resources. We've heard nothing from Johnson's crew on this (although the Tories have been traditionally very reluctant to introduce meaningful constraints on dodgy funds flows and Johnson cares even less about the rule of law), but John McDonnell has tried to assure people that Labour won't introduce capital controls while talking about plans that effectively require them. Problem is that governments may not warn people before introducing capital controls for obvious reasons...;

• breakdown in the rule of law - traditionally not associated with the UK, but that reputation has been vastly undermined by the antics of the Conservative Party during the past four years and particularly the Johnson crew during the Leave campaign and over the past few weeks. Threats not to implement customs checks or to withhold payments due to the EU under budgetary commitments are particularly concerning. Nobody wants a deal that involves constant disputes and endless international arbitration proceedings; and businesses won't trade in violation of terms where that lands them in hot water or isn't otherwise in their interests.

Clearly, any government also has to be careful not to lose sight of many other factors that contribute to people and businesses' assessment about the relative attractiveness of doing business in a country more generally - such as unusual or extraordinary taxes on income, property and transactions.

The Latest Designer Drug: 1D10C

Drug classes A, B and C to remain

UK drug enforcement officials say a new drug has taken hold in Britain, pushing 'traditional' class A drugs aside.

Code-named "1D10C",  the new drug comes in powdered and liquid form, and can be administered through any external body orifice, with many even preferring to receive it in the form of inhalers, eye drops and ear drops, making detection almost impossible.

Effects include unexplained euphoria, self-confidence and loquaciousness - the tendency to be overly talkative.  

Side effects include paranoia, delusional episodes, lack of concentration, the inability to grasp complex problems, lack of empathy and poor verbal and non-verbal reasoning.

While not yet classified, officials have begun investigating the sources and usage patterns of 1D10C in Britain. Recent admissions of drug usage among political candidates has drawn officers' attention to their ranks, as well as individuals working in media, public relations and political lobbying organisations, as well as large political donors.

While declining to give further details or to be named, one senior investigator said, "We see the flow of political donations in support of projects that are obviously flawed or based on a mistaken understanding of how certain processes or industries work as likely indicators that this drug may be being distributed and abused in large quantities." She declined to say when the investigation would end, or the likely outcome, but one possible result would be a recommendation that the drug be classified in order to restrict its availability.

Both the Conservative Party and the Labour Party declined the opportunity to comment.