What a world we live in, not only is there the crime of fraud, there is now the option to automate it. Yes, the Guardian is giving us (at https://www.theguardian.com/technology/2017/aug/08/fake-news-full-fact-software-immune-system-journalism-soros-omidyar), the automation against ‘fake news’. It is the subtitle that gets the blackberry pie in this. With ‘Full Fact software backed by George Soros and Pierre Omidyar fact-checks statements in parliament and news media in real time‘ we see the start of a series of events, and I am honestly not sure where they are trying to take it at present. The bias is intentional, because we all have this feeling, when a billionaire (or plural) support something, it will put them in a better position. It could be seen as unfair bias, yet bias is in the human core, which in light of what the media does more and more not the worst position to place yourself in.
To get there, we need to take a few steps. With fraud we need to see where that is at. We get “wrongful or criminal deception intended to result in financial or personal gain“, so let’s take a gander into The Fraud Act 2006. This gives us three options.
The first being that with fraud by false representation we can go hilariously with ‘I really have an 11 inch dick‘, or better stated, in section 2 paragraph 2 it is all about:
A representation is false if: (a)it is untrue or misleading, and (b)the person making it knows that it is, or might be, untrue or misleading.
This issue as seen in the first part as it could be measured and acted on by the rulering (pun, innuendo and Full Fact challenge intended), yet the statement ‘I am the greatest lover in the world‘ is that fraud, being delusional of conviction of a personal ability through (or enabled by) the deceit of your own ego, so how to prove such subjectivity, against fraud? Now the important part here, which is not covered in lust (or perhaps it is) is seen in paragraph 3 with
(3) “Representation” means any representation as to fact or law, including a representation as to the state of mind of (a) the person making the representation, or (b) any other person.
and the important part that matters in his case is paragraph 5 which gives us “(5) For the purposes of this section a representation may be regarded as made if it (or anything implying it) is submitted in any form to any system or device designed to receive, convey or respond to communications (with or without human intervention)“, this puts tweets, social media and optionally this software solution in a similar weird new position.
Now this is part one and you can see where this is going, and there is no twist to this part of the tale, yet the other two are still important.
The second is fraud by failing to disclose information. Now in this case journalists get a free pass, because it is about the legal duty and that person is in breach of this section if he (or she)
(a)dishonestly fails to disclose to another person information which he is under a legal duty to disclose, and (b)intends, by failing to disclose the information
- (i)to make a gain for himself or another, or
- (ii)to cause loss to another or to expose another to a risk of loss.
so this tends to be professionally set to the makers of laws, accountants, those people with red or purple robes and wigs, you know the types, in this case doctors, and not to forget certain contractors working for governments, in this the military and intelligence community cannot be convicted or prosecuted as they have other sets of rules, like national security, military law and in some cases maritime law could leave a person not in breach of this, yet they would have their own set of rules that still sets clearly their responsibilities.
And last we get fraud by abuse of position. I am merely mentioning this almost for the completeness of the fraud as it is in common law. Yet with
(1) A person is in breach of this section if he (or she), (a) occupies a position in which he is expected to safeguard, or not to act against, the financial interests of another person,(b) dishonestly abuses that position, and (c)intends, by means of the abuse of that position
- (i)to make a gain for himself or another, or
- (ii)to cause loss to another or to expose another to a risk of loss.
(2) A person may be regarded as having abused his position even though his conduct consisted of an omission rather than an act.
Actually, now that I think of it, the makers of the software program once called the “bullshit detector” in an early version of the system by its creators, is now set to be a virtual truth vaccine. I am in two set mind here, on one side I love the idea and on the other side there are a few hitches that would make it fall over for possibly a lot longer than anyone could consider. You see, when I think of fake news, I do not think merely of the actual fact misusers. The joke that Channel Nine became in the evening as they got the first lead on a shooting (at https://www.youtube.com/watch?v=JjkQSREjxH0), the lady now nationally known as the ‘chick chick boom girl’, gave an eye witness report and the journo just ate it all up. A sexy looking girl giving the news all the ‘facts‘ as she was allowed to report this in the limelight offered. The reporter felt so good and six days later the Age gives us “Within 24 hours of the footage first airing on Channel Nine, T-shirts emblazoned with “chk chk boom“, the phrase which Ms Werbeloff used to describe the gunshot, were being sold on a US website. There were also drink bottles with the slogan and a rap-music remix of her video, and in addition, the fact that the news was bogus, fake, not real and pretty well made up“, we now have two issues, she is not merely guilty of fraud, she possibly impeded an investigation into a shooting.
This is more than a funny story, even as we can admit that one comment in the YouTube states that it is still funny and to some extent I agree, merely because the gullible reporter basically got played. We can set in equal measure giggle to the smallest extent as KVTU reported that the names of the MH370 crew members were Captain Sum Ting Wong, Wi Tu Lo, Ho Lee Fuk and Bang Ding Ow. The fact that the newsreader and the editor did not ‘seemingly realise‘ that this was fake is a failure on an entirely different level and no software on the planet will protect us from stupid people (or gullible politicians for that matter). So that in the back of my mind within a minute of the article in the Guardian had me pondering this ‘software solution‘. Now, I do believe that the software could be used to flag thousands of messages almost instantly, yet the claim we get from “The early version of the software scans the subtitles of live news programmes, broadcasts of parliament, the Hansard parliamentary record, and articles published by newspapers. It tracks millions of words sentence by sentence until it identifies a claim that appears to match a fact-check already in its database” is dangerous on a few levels, yet as the makers interestingly and validly point out: “Babakar is keen to stress the limitations of the system so far and believes the tool should only be used by journalists in the first instance rather than the general public“, I like that part, they admit that the software is nowhere near ready even when it initially launches.
So where is my issue?
There are a few sides, with ‘claim that appears to match a fact-check already in its database‘ we have to realise that the vetting process is a critical part in this, so how soon will we see clearances and cleared checks of miscommunications, the previous news cast from KVTU being a first example. This process would require internal political hands from whoever uses it and there is where the bias sets in. We might go biased and auto accept the facts from sources like the Washington Post, the Times, the Dutch NRC Handelsblad and the Swedish Dagens Nyheter, yet now we get the second part, we have all seen and under the best of conditions there is the danger of getting ‘lost in translation‘ (not the movie which was brilliant). So when we apply the lost in translation, we have the local versions (UK version of English) with the Sun giving us today: ‘1,200 killed by mental patients‘, so is this fake news? I would state so, but knowing the people behind the Sun, they will have some numbers that add up to 1200. Yet ‘today’ was October 2013, when it actually was ‘today’. So here we see the first application of lost in translation, the second side in that headline was seen when we go to the ‘actual’ news, were we see “It discovered 1,216 people were killed by patients with mental illness from 2001-2010 — an average of 122 deaths a year” (at https://www.thesun.co.uk/archives/news/1052064/1200-killed-by-mental-patients/), so is this still fake news? That same article gives us “The study — the first of its kind in the UK — found 45 per cent of those with severe psychiatric problems were victims of crime in the previous year“, now giving weight to anything the Sun states is one thing, accepting the good looks of the lady on page three is another, yet in the end, were these stated ‘facts’ lies? I have had loads of issues with tabloids for the longest of times. Still, when we now look at fraud, where ‘A representation is false ifit is untrue or misleading, and the person making it knows that it is, or might be, untrue or misleading.‘ The additional ‘a representation may be regarded as made if it is submitted in any form to any system or device designed to receive, convey or respond to communications (with or without human intervention)‘. Here is the initial ballgame, you see, it is not merely about fake news, it now becomes about the interpretation of language. In its most basic terms, a grammar system designed to comprehend innuendo. When we know that loads of Journalists are incapable of comprehending sarcasm under even the most optimal conditions and in addition, when I state that sarcasm that backfires is merely irony, they are likely to become even more confused. So the heavens might love the simple minded, but here it becomes an issue on a much larger scale.
The article gives us “The fledgling system is not without its problems; sometimes it flags up a fact-check that isn’t relevant, for example. The challenge for the programmers is to get the software to understand the fuzzy logic and idiom used so often in speech” This actually barely scratches the surface. You see, there are a few other parts, how would the Metro front page ‘Brexit begins‘ be seen? Officially, the letter was delivered, yet the overwhelming ‘news’ is that Brexit has not started yet, could flag this news article. The truth is the interpretation of what constitutes the start of Brexit is also the issue. It’s not like having sex (yes the example matters, so do read on), that is the act and point in time, it is what happens after. So when is the baby seen as alive? When the conception has happened? When the child is capable of being born alive? In Australia it is seen after 28 weeks of pregnancy. Yet in separate Australian states, in this case Tasmania, Victoria and Western Australia, the terminal points for abortions are 16, 24 and 20 weeks, so there is an issue on a few levels. In cases there is the Crimes Act 1900 (sections 82-84), whilst the ACT had abortions repealed as a criminal offense from the Crimes (Abolition of Offence of Abortion) Act 2002. So, one moment in time that on a national level already has 4 different assessment dates on one national calendar. In that mess on one ‘simple’ issue, we see a mess that becomes even messier when theology gets in the way of the subject of conversation and that is more than just a reference to separation of church and state. This shows part of the problem, also for the us where federal law can be a real mess for the non-legal people and that also has implications for the ‘fake news’ vetting system. So how could any article on illegal abortion be weighed correctly, especially when a politically or religiously tainted tabloid becomes the provider of that news?
Now, if they pull it off would be quite a feat, yet when we go back to the initial fraud part, if the system flags an article as possible fake, what happens when the news agency in light of thousands of items a day scraps that one item? Would the non-reported article by the journalist be seen as a matter for an aggrieved status? That is part of the problem these makers face, because in some nations the torts law is pretty strong and if there is a case of ‘redress of damage‘ the roll of the dice would be well worth it. There would be a direct and instant case of insurance companies, especially in any editorial status to not warrant any level of insurance when such software is used. Not now, possibly even not ever, although the latter part would be highly speculative by me and I hope to some extent that I am wrong, because pulling it off, getting software to be this able would be quite the achievement and the essential requirement for any evolving AI. This is because intelligence artificial or not would be about the ability to weigh information and facts to the proper value of innuendo and ascertainment of paraphrased facts. Weighing information is a skill that is almost never done correctly on the present day internet. A nice example is to compare the search engines Bing and Google, when you seek ‘UK torts‘ the search engine Bing gave me as a 4th and several subsequent mentions ‘Tort law in Australia – Wikipedia‘ as well as a few other Australian references to Australian law. Was that what I had asked for? No, it was not (which is why Bing is my preferred source for not finding facts, yay Google).
All issues given that could be seen as fake news, or is that fake information, so when we see the flags, how long until the bias sets in? You see that is my largest issue with what could come, with bias I get what they want me to see, which I touched on in the very beginning, because automated or not, censoring is a dangerous principle under the most ideal situations. With software, no matter how good the intention we could see a dangerous curve towards what is automated assumed we wanted to see. You see, there are too many parts of the dimension of ‘fake news’. This has been the larger issue with media for a much longer time. In this, Mevan Babakar the project manager at Full Fact in London has his work cut out for him, I do not envy him, yet if this gets solved and if it works, the value of this software would be a lot larger than most people could realise, and it would be a technological software marvel to behold.