Delivering Libellous Content

March 17, 2014

The DeliveryDemon had to chuckle at this article

The law has certainly been working hard to catch up with technology, and the impact of this sort of libel is very real to those who are libelled. But the legal profession is missing a trick here. Behind the scenes, there is technology which looks for keywords and tries to interpret them. By and large this software is still remarkably primitive. It has yet to get to grips with the ability to interpret the context. Basically it lacks ‘intelligence’. It is designed to provide an answer at the expense of providing a sensible answer.

Google predictive text gives some good examples of what can happen and various mobile phone predictive text engines can be even funnier. The automated parsers used by recruiters cannot distinguish between Coral the bookmaker and Coral the programming language. Amazon’s ‘you might like’ suggestions suggest you buy an identical item to a recent purchase, with a different brand name.

To some extent, many of these tools are designed to depend on data which is not quality-controlled in any effective way. Certainly an Amazon vendor will enter the keywords likely to maximise search hits. that can mean the entry of keywords with little relation to the product being sold.

Google is one of the more sophisticated players since its product depends on understanding what a searcher is likely to want, but the Telegraph article shows how primitive the logic is. Asking users to log in and relating searches to their search history has the potential to improve search result quality, but people are becoming increasingly sensitive to the amount of their data held by large corporations, and legislators are starting to respond to those concerns, so relying on users logging in may not be the most fruitful development path for this type of tool.

The examples in libel article certainly have merit. Either the tool is not fit for purpose, or it is being used unintelligently. A fairly obvious solution would be for the news website to flag articles as being either positive or adverse, provided the tool refrains from coming up with links to ‘similar’ articles unless they were also flagged as adverse. If the tool can do this, the web publisher is at fault. If the tool can’t do it, then there are two potential breaches. The tool may be inadequate for the purpose for which it is being sold. Or the web publisher may be making inappropriate use of the tool. Of course, when a payment model is based on click throughs, the incentives tend not to favour anything which limits the number of links displayed.

A fruitful approach for legislators would be to look beyond individual libels and examine the capabilities of current tools, and the processes which web publishers use to to mitigate the risks arising from tool limitations.

There’s More to Risk Than Probability

June 3, 2009

The DeliveryDemon is a Management of Risk (MoR) Practitioner so she is used to thinking of risk in terms of probability, impact and proximity:

  • How likely is it?
  • What’s the worst case scenario if it happens?
  • Is it likely to come to pass soon or in the distant future?

June’s issue of Project Manager Today contains an article by David Hillson on risk perception, which refers to work by risk psychology researcher Dr Paul Slovic on factors which influence perceptions of risk. DeliveryDemon thinks that Dr Slovic’s work should be considered in the light of existing well-established MoR tools for identifiying and managing risk.

‘Dread If the outcome of a particular risk is something we imagine to be terrible…., our perception of the risk is heightened.’ Simply not true. The outcome (impact) is part and parcel of the risk, not something separate. The worse the potential outcome, the greater the risk. An illness with a 40% fatality rate is 100% fatal to the unfortunate 40%. The dread is associated with that harsh 100%, which is a factual numerical component of the risk, not just a perception.

‘Control When we believe we have control we perceive the risk as lower.’ When we have control we have tools to manage risk. Different people make different choices because tolerance of risk varies from person to person, and from organisation to organisation. If we have control we may choose, at one extreme to walk away from the risk, or at the other extreme, to live with it. We may take action to reduce the risk, or to pass the risk on to someone else. The decision may be not to drive that Porsche through the icy hairpin bends of a mountain pass, or to do it anyway and live with the consequences. The risk may be reduced with snow chains, or passed (partially) to someone else through an insurance policy with a good vehicle replacement clause. None of these options are available to the trussed up prisoner of a car thief whose drug intake has sent his risk tolerance sky high. Control allows the individual to match risk and risk tolerance. There is logic to this, not just perception.

‘Natural vs Man-Made Hazards resulting from human actions are seen as more risky than natural hazards.’  Control comes into this again. The geographical distribution of natural hazards is well known and people may choose not to live on the San Andreas Fault or the slopes of Stromboli. If brought up in a danger area and tied to it by family commitments, people know the warning signs and have learned ways of predicting and mitigating the problems. These days, people who find the risk level too great often have the option of moving away. Man-made hazards follow a different timescale. The announcement of a planned man-made hazards affects a previously blight free area, and individual inhabitants have little say over the introduction of the risk or its management, so have no tools available to match the risk to their risk tolerance.

‘Choice If I have some choice over my exposure to a risk then it seems lower than if I am exposed involuntarily.’  This comes back to control. If I have choices I am able to mitigate the risk by reducing the probability, the impact, the proximity. I can match the risk to my risk tolerance. Again, this is not merely a matter of perception. Choice allows us to alter numerical risk factors.

‘Children Any risk which affects children is seen as worse than one which affects only adults.’ Possibly. There are cultures which venerate childhood, and cultures which view children as more replaceable than their parents. DeliveryDemon is not an anthropologist so isn’t going to comment on this one.

‘Novelty New risks are seen as being higher than ones we have grown used to seeing’. A variation of ‘familiarity breeds contempt’. This is an extremely simplistic view. When a risk is new, less knowledge is available. The cautious will weight their risk assessment pessimistically to allow for that lack of knowledge. However, those with a stake in the risk source may equally discount the risk level on the basis that there is, as yet, no evidence to substantiate the risk, a common statement during discussions of the risks of large-scale GM planting. The Novelty factor may work either way. Similarly, lack of novelty can work to increase or decrease actual risk. That the building industry has a has a high level of injury and fatality may be attributable to a macho attitude to risk which leads to the cautious being viewed as wimps. The culture creates an additional risk which needs to be managed. In the abstract, extreme sports offer a higher level of risk than a building site, but the culture is very different. Conventions and support mechanisms have evolved such as diver buddying, parachute checking, frostbite checking, mountain rescue. There is an interest in the development of the individual and of technology to enable boundaries to be pushed. Participants court a dangerous edge, but they also seek the risk mitigation tools which will allow them to push beyond current boundaries.

‘Publicity If a risk has a high profile in the media… it will be perceived as being more risky.’ This is all about understanding the impact. A dry paragraph mentioning an explosion in a distant city leaves it to the imagination to guess the impact, whereas a picture of mangled bodies leaves little room for doubt. The fact of publicity has less impact on risk perception than does the presentation of the published content. The theory behind this underpins the entire advertising industry. At the other extreme it can be used to explain mass hysteria. Publicity is less of a risk modifier than a risk in its own right. For example, late last summer there was a risk of severe economic downturn. Publicity helped it on its way by making people fear for the future to the extent that consumer spending was drastically reduced. It was less the fact of economic threat than the way it was presented in the media which prompted that. For example, the BBC might be expected to provide a dispassionate view of the situation, but Robert Peston’s highly emotive reporting style was a significant influence on consumer decisions to cut back on spending.

‘Propinquity If I could be a victim the importance of the risk seems higher than it really is.’  The word ‘seems’ is inappropriate here, as is ‘importance’. Impact is as important as probability when assessing a risk. If I am assessing a risk from a personal perspective, the 10% probability of occurrence is only a small part of the story if the impact is 100% fatality. The risk is genuinely greater, it is not just perception.

‘Risk – Benefit Trade-off If exposure to a risk could also result in a perceived benefit as well as a threat, the risk is discounted.’ Business case, anyone? It’s not actually a discounting of risk, more a balancing. It can be done on the basis of gut feel or total logic or a mixture of the two. Benefits may blind people to the risks but not necessarily. The DeliveryDemon knows that mountains and mountain weather can be dangerous, and has a lot of experience of assessing the risk. That doesn’t stop her heading out in (nearly) all weathers in search of benefits such as air like wine, panoramic views, solitude, wildlife, crisp snowfields and icewalls and glaciers, a good downhill run, or even a blast of wind and weather. The balancing bit of risk management is the prior selection of equipment, assessment of fitness, and choice of route. It’s not discounting of risk, it’s a balancing act.

‘Trust Where protection from risk is offered by a trusted party the risk is perceived as lower.’  In MoR terms this is transfer of risk, and the insurance industry is a classic example of it. It works by reducing the actual numerical risk at the impact level. If I’m travelling, insurance doesn’t reduce the probability of my luggage going missing, but it does reduce the impact because I can reclaim the cost from the insurance company.

The DeliveryDemon finds Dr Slovic’s classifications quite naive and much prefers MoR’s simple, factual and pragmatic analysis as a tool for programme and project managers.

Seriously Extreme Project Management

April 24, 2009

Think of a project with all the odds stacked against it:

  • High profile
  • Highly competitive situation
  • Complex logistics
  • Short and inflexible time span
  • Small team of strong minded individualistic experts
  • Extremely high level of risk
  • Absolutely zero tolerance when it comes to simultaneous delivery of all 6 elements of the project

Photographer and elite athlete Tony Lamiche had 5 days to deliver a single photograph showing all 6 major mountain sports in action:

  • Climbing
  • Speedriding
  • BASE jumping
  • Skiing
  • Paragliding
  • Snowboarding

Look at for the most impressive example of project delivery the DeliveryDemon has ever seen.

Surprise, surprise – Risk Management

April 22, 2009

The DeliveryDemon was interested to see some risk management being discussed in the US Treasury.

The US bank bailout includes the setting up of a public-private partnership to buy up toxic assets. As this Reuters article explains:

  • The cost risk to taxpayers outweighs the potential for benefits
  • Conflicts of interest have been identified and recognised as a source of risk
  • The scheme is inherently vulnerable to fraud, money laundering and other forms of abuse
  • The ‘public’ element of the public-private partnership dilutes the risk for the private element, increasing the likelihood of a high risk approach to managing the overall funds

What is surprising is that these risks are being so publicly and simply stated. The early days of most public-private partnerships are normally wreathed in a mist of bonhomie as each party strives to protect and enhance its relationship with the other party. Reservations are rarely made public.

The DeliveryDemon will be interested to see how strongly this risk management and transparency is followed through. And whether the UK takes a lead from the US when it comes to acknowledging and managing the risks associated with the UK government’s bank bailouts.