This session examined the role of automated translation technologies combining statistical and artificial intelligence approaches in servicing the future needs of a multilingual internet. Discussion also referenced the increased incidence of commercial filtering, personalisation and behavioural targeting to individually tailor services and content options for Internet users. Consideration was given to the role of the Internet in eroding the costs of accessing diverse sources of information, alongside concern that excessive reliance on filtered and personalised services/content can undermine creativity, spontaneity and choice. Future education models will need to focus more on how to authenticate and exploit online information rather than traditional approaches of learning though memorisation. The low cost production and rapid prototyping opportunities offered by 3D printing technologies will revolutionise and disrupt the global manufacturing industry, although many of these innovative uses are likely run into conflict with existing copyright and IPR regulations. Attempts by governments to apply national laws to the Internet in relation to inflammatory or defamatory material will potentially pave the way for further regulation of other aspects of online activity which runs the risk of creating an increasingly balkanised Internet. Alongside the need for digital preservation strategies to address the long term sustainability of physical and online information, a key challenge for Internet governance will be to strike a balance between vertical information management structures and effectively harnessing the unprecedented potential of the internet for horizontal communication and engagement.
(links take you to the relevant forum subtopic):
Q1: In 10 years will the concept of a free and open Internet be an anachronism?
Link to main discussion topic:
Use of automated translation can be problematic in diplomatically sensitive circumstances. In some instances the use of free but imperfect or inaccurate tools can have negative outcomes. Frequent problems are encountered in processing acronyms and industry or sector specific terms as well as vernacular/colloquial expressions and jokes. It was conceded that in more informal settings even imperfect automated translation still has the capacity to provide valuable/helpful levels of access to multilingual resources.
In 5-10 years there will be automated translation techniques which will adequately support 90% of most communication needs. This will involve a combination of statistical and artificial intelligence methods along with increased mobilisation of crowd-sourced input. India in particular is potentially fertile ground for supporting advances in this area given the number of official languages in that country alongside high level intellectual and engineering resources. In general terms it was predicted that within a decade there will be more and better automated translation services which will be serving more people across a greater range of languages.
Artificial intelligence techniques are already being used by the intelligence community to analyse large numbers of articles, blogs and websites in order to assess moods and trends. The next step will be for these kinds of approaches to be rolled out with increasing frequency in the consumer sector (which is already the case to an extent in terms of online product recommendations and behavioural advertising) to generate valuable data for suppliers of commercial products and services. This will increasingly be about second guessing what options particular users want and providing them automatically. As users become increasingly at home with such services their behaviour and approach to using information and technology will be accordingly altered.
Some advanced that this will be a positive time saving development – and has significant implications in relation to new approaches to education. Is it relevant to teach children to memorise key facts or statistics that they can easily find out through a search engine? In a context where information is progressively easier to identify and acquire, education approaches should focus more on how to authenticate and exploit this information rather than committing it to memory.
There is a distinction between finding the information you need and finding the information that commercial entities have persuaded you that you need in order to maintain their revenue streams. While the use of automated filtering and personal tailoring of information and services is valuable in helping us navigate the vast and expanding universe of available content and services – the importance of choice should not be ignored. Choice is a central concept within economics, and where the availability of choice is constrained or reduced then arguably freedom is constrained or reduced as well. These restrictions may serve to undermine opportunities for spontaneous or creative discovery of ideas and concepts which exists outside our day to day terms of reference (serendipitous discovery). It was advanced that citizens need to be educated and provided with tools which effectively constitute sharp needles they can use to puncture their filter bubbles.
The concept of the online filter bubble was substantially derived from the work of American legal scholar Cass Sunstein. Sunstein argued that individuals with certain political and ideological leanings tended to consume media and information which confirms their preconceived beliefs and inclinations (Democrats watching CNN and Republicans watching Fox News). For example it could be argued that if reading your daily Twitter feed (news feeds from Twitter users you are following) doesn’t regularly make you irate or upset then you are probably not following a sufficiently diverse cross-section of society.
In contrast it was suggested that the Internet itself should be seen as the sharp needle which has punctured the historical information filter bubble. During the immediate pre-Internet age and television represented the primary source of news and information (with newspapers and radio operating as more distant secondary sources). With the arrival of the Internet this limited range of information choices has exponentially expanded beyond recognition delivering a spectrum of social, political and cultural content which simply would never have been accessible in the pre-digital age. The presumption that people are only exposed to online content and opinions they already agree with is mistaken. On social networks people are bombarded from all sides by newsfeeds, posts, blogs and links from more diverse sources than they have ever previously encountered.
Much of our previous appetite for content diversity was constrained by the costs of accessing information. An individual might be interested in a range of different areas but would lack the sufficient resources to acquire content or information across all those areas of interest. With the arrival of the Internet this cost barrier disappears which empowers and enables individuals to radically expand the horizons of their consumption of news, content and information. Optimists would argue that information technology drives more (as opposed to less) opportunities for diversity of experience.
Being overwhelmed by information is a developed world consumer problem which information technology companies will spend significant resources trying to solve given the significant opportunities for profit in this equation. It was also suggested that there is a generational component to this – whereby older generations are more likely to feel overwhelmed by the myriad of different potential sources of information and content, the young are more likely to metabolize these developments by employing new tools and techniques to effectively manage these information sources. It was predicted that private companies (both large corporations and start-ups) will invest significant amounts of money to develop services which will relieve the human brain of having to deal with functions and activities which are menial enough that they could be covered equally as well or better by a machine.
People increasingly rely upon social networks (e.g. Facebook and Twitter) and professional networks (LinkedIn) to filter, recommend and suggest information they might be interested in. This process also plays a part in undermining traditional broadcast news delivery channels and promoting more flexible on-demand style consumption of information and media.
It was predicted that the adoption of 3D printing technologies will revolutionise manufacturing in Africa and reduce reliance on Indian and Chinese imports. The most disruptive technologies are usually those which are cheap enough to be adopted and used by large numbers of people. 3D printing technology will become increasingly low cost and support cheap and rapid prototyping processes which enable people to swiftly turn ideas into physical tools and products. 3D printers currently produce solid thermoplastic objects, but this could soon expand to cover metal and ceramic objects. Libraries could be key community providers of access to this new technology.
One of the biggest looming issues surrounding 3D printing is the shadow of oppressive international copyright regimes and the likelihood that many innovative and useful functions (particularly in the developing world) will be considered violations of intellectual property rights.
One of the chief sources of China’s economic competitive advantage is not simply low cost labour (many countries offer labour at lower cost) but their capacity to offer rapid prototyping of products. One of the key economic benefits of 3D printing will be to enable many other developing countries to offer this facility.
It was also suggested that low cost 3D printing could also support a greater propensity among consumers to repair products, devices and appliances (if small but crucial replacement parts can be quickly and easily manufactured). However this would depend upon the availability of the digital designs/data files which could support such activity – which may of course be restrained by copyright and IPR regulations.
The capacity of the Internet to connect people and businesses across national jurisdictions has begun to steadily erode the power and relevance of the traditional nation state. However, it would seem that few of these nation states have fully woken up to that reality. As governments start to recognise this development, it is probable that they will react by attempting to divide or segment the Internet on a national basis – which will lead to an increasingly balkanised Internet experience. In many ways historically speaking the Internet has gotten a “free pass” from governments for longer than might have been expected.
Right now there is rising concern on the part of governments seeking to apply their national laws to a global borderless Internet. Ranging from the prohibition of Nazi propaganda in Germany to the prohibition of anti-monarchy statements and expression in Thailand – many countries proscribe specific types of behaviour within the boundaries of their nation state which the Internet makes progressively more difficult to effectively enforce. Sites like Twitter and YouTube receive daily court orders in different countries asking them to remove material which is considered defamatory or inflammatory in a particular national jurisdiction (ranging from hate speech to jokes, political criticism or statements which are deemed to be unacceptable or perceived to undermine public morals). Many of these laws (such as the hate speech laws in France and Germany) were enacted using perfectly sound democratic processes and principles. This domain effectively represents the thin end of the wedge where governments will set publically acceptable precedents for regulating Internet content – precedents which are then likely to pave the way for further expansion of the scope and reach of this regulation to other areas of content and Internet activity.
It will become increasingly important to address the long term sustainability of both physical and digital information and ensure that this knowledge is diffused across society. Traditionally libraries have served as the archivists of information and knowledge – potentially this approach should be applied to the digital world. Without a comprehensive approach to digital preservation, physical books may be lost forever while inane Facebook comments will be preserved for eternity. We currently create more new digital information each day than was created in during all of human history. If physical books are not available online they might as well not exist. Historically speaking less than 1% of human knowledge was captured in physical books, which makes those that remain even more valuable given that vast amount of information and knowledge that we have already failed to capture/acquire.
A balance needs to be struck between implementing top down centralised information management structures/regulation and harnessing the capacity of this Internet for horizontal communication and engagement whilst drawing upon the innovative potential of crowd generated information and knowledge.