Tag Archives: Artificial Intelligence

The issue of personalisation and its impact on KM

Author Hal Kirkwood BIR Board member, Bodleian Business Librarian, Said Business School, University of Oxford. President, Special Lib Assoc. 2019

The current state of affairs was on full display at the last (November 2018) KM World Conference in Washington DC.  I had the opportunity to attend for several days to see first-hand what is happening in the knowledge management realm.  There were many themes prevalent throughout the conference; each day consisted of 3 tracks. The Day One tracks focused on KM & Culture, Digital Workspaces, KM Tools & Tech.  Day Two tracks focused on Knowledge-Sharing Processes, Content Management, and KM Culture & Collaboration.  Key takeaways and themes were the importance of collaboration; identifying the right tools to fit the problem and your organization’s culture; designing environments, both physical and virtual, for employees and clients; determining how to transfer knowledge; developing information ecosystems; and the implementation and impact of artificial intelligence and machine learning.  The clear underlying theme is the continuing intersection of people and technology.

One aspect that is gaining traction into KM is personalization; utilizing individual user data to provide a more focused recommendation or timely suggestion. Technology, in conjunction with access to massive amounts of data, is driving momentum towards ever greater personalization.  Personalization, not customization.  Consumers become weary of making choices when these systems can make relevant choices for them based on their prior experiences.  Consumers are showing preferences towards companies that provide effective, relevant personalization.  However, since knowledge management focuses on the internal management of a company’s knowledge personalization at the employee level has been slower to develop.

Personalization has primarily been within the purview of marketing and consumer buying habits.  The power of personalization relies on a combination of data that was once inaccessible; namely geolocation crossed with purchasing habits.  It has become especially powerful when the immediacy of time is included to deliver personalized information and recommendations to a potential customer at the most optimal moment to affect their behavior.  Artificial intelligence and machine learning will make significant inroads in the personalization strategies of companies marketing plans to provide more focused experiences for customers.  1

The challenge for many companies is to scale this personalization to the masses.  AI and machine learning will increase the capacity to track multiple data points for larger numbers of customers. This will increase the expectation of customers for improving levels of service that meet their exact needs and requirements.  Evidence shows that it is highly successful when implemented in increasing sales and customer satisfaction. but that most companies are not implementing it.

Every company is now looking for ways to gather customer data that can be used to make more informed, and more specific, decisions on individuals.  Many companies are also capturing terabytes of data on customer behavior to then sell to businesses for this very reason. There is the issue however, that the attempt at personalization will be wrong based on the AI processing poor or inaccurate information.  As personalization becomes more accurate, and more ubiquitous, it will seem all the more glaring when AI-driven personalization is incorrect. Consumers are likely to feel more uncomfortable about what data is ‘out there’ on them and its accuracy, or lack thereof.  This is a complicated issue of human perception of technologically driven services.  How much control we have over all of this data is also a major concern.  In Europe, GDPR is beginning to make an impact by providing consumers with more control over what data is collected and how it is used.  It remains to be seen how exactly this will impact the data collection and utilization process. Many consumers, when surveyed, approve of the use of their data if they will receive a tangible benefit. There are some conversations taking place about implementing some form of GDPR in the United States, but little in the way of concrete details have provided.

Companies such as Netflix, Spotify, Amazon, and several other key companies are pursuing, and leading, the development of even greater data collection to develop ever more enhanced services for individuals.  Areas like physical fitness, healthcare, and personal finance are becoming driven by apps that collect personal data to then provide recommendations relevant to an individual’s life.  Consumers will allow themselves to be tracked in this way because of the return on investment of their personal data.

The majority of personalization development has been in the B2C marketplace; there will likely be increased demand for it on the B2B side.  The key element will be systems that collect client-level data that can be assessed by AI applications.  Many companies are moving into this to deliver solutions for collecting and analyzing data.  Business intelligence systems will develop as AI and machine learning are layered into them for much greater personalization of services and deliveries to corporate clientele. Companies must make the choice to implement an AI-based system to drive their decisions.  Not an easy task when it often requires a significant operational and cultural shift in how they conduct business.  Companies making this decision are likely to benefit but must be wary of the myriad pitfalls.  What ramifications this will have on the competitiveness of companies and markets, as well as within the broader business information environment still remains to be seen.

Emerging skills for the information profession – The 4th theme in the BIR Annual Survey

Over the past two years Business Information Review has examined a range of emerging technologies that are beginning to impact on professional practice in the commercial information management sector. These have included smart technology, cybersecurity, Augmented Reality, and Virtual Reality. We have also explore a range of social and regulatory issues associated with emerging technology including GDPR and fake news. The information profession has become closely aligned to technological change, and information professionals have often been early adopters of new ways of communicating, managing, and finding information, data and resources.

The issue that has recurred most frequently over that time, both in the journal itself, and in the conversations that we have with professionals to discuss which professional trends the journal should be addressing, has been the growing place of Artificial Intelligence (AI). AI poses advantages as a tool in information management, but also challenges as a disruptive technology for the profession, business services, society more widely. AI has featured as a topic in the journal in March 2018, December 2017 and March 2017. And it is featured again as a dominant theme in the 2018 BIR Annual Survey, but in two ways.

In July this blog reflected on the ways in which AI is poised to transform information work and business processes. But that change and other associated technological developments pose a different set of challenges for information professionals, implying new ways of working, and an associated new set of skills and knowledge. The final theme in this year’s BIR Annual Survey reflects the ways in which senior information and knowledge professionals in the commercial sector are beginning to tackle these challenges, and confront the changing skills-set of the future information professional.

The BIR Annual Survey is the longest running continuous survey of the needs and working lives of commercial information and knowledge managers in the World. Since 1990 it has provided an invaluable insight into the changing world of Information and Knowledge Management. We like to think of it as an annual snapshot of the state of the profession. Throughout July and August we have provided a taste of the issues that are preoccupying information and knowledge professionals in the 2018 BIR Annual Survey. The final report will be published in the September issue of Business Information Review, and provides a fascinating insight into a rapidly changing profession.

On hollowing out….

Author: Stephen Phillips, Executive Director Morgan Stanley and BIS Editorial Board Member

Please note this post contains the personal views of the author and are not connected with his employer.

Earlier this year Stephen Dale wrote a fascinating article on corporate memory for the May edition: “Are we destined to forget everything we already know”.  As I reflected on his narrative, I felt the need to explore this topic further, as organisations appear to have become “hollowed out” as they focus on cost to deliver short-term efficiency and opportunity.

I also felt the need to re-interpret some of the terminology used to define information, knowledge and memory.  The vocabulary for these concepts has become interchangeable in many organisations as they continue to search for increasingly challenging opportunities to realise further benefits from managing this space.

A quick search on Google (I know!) reveals the first definition of knowledge to be “facts, information, and skills acquired through experience or education; the theoretical or practical understanding of a subject”. Nothing contentious there, but the second definition cites it as “information held on a computer system”. The latter was a new one to me; since when did knowledge become defined as information held on computer systems?

Another interpretation rang more true to me: “awareness or familiarity gained by experience of a fact or situation”.  To my mind, this speaks to the human nature of knowledge – it is much more than facts and information; it is about awareness, familiarity, experience, consciousness, perception and appreciation.  All nouns that reflect human nature and remain technological aspirations; at least for the time being.

Whilst it is important to recognise and appreciate the capabilities of the latest developments in AI, machine learning and neural processing, it is more important to recognise their limitations and appreciate the benefits associated with tenured people and their accumulated know how in their respective roles.

The most impactful force in the resizing of the business information industry has been the empowerment of “knowledge workers” to do their own information seeking.  However, investment in these workers and their information skills has lagged behind, leaving a workforce that know which buttons to press but who are poorly informed about what underpins the information and technologies they use every day.

Redundancies, outsourcing or offshoring of business information specialists compounds the issues.  New entrants that come into the industry find it difficult to secure positions with their limited experience which is incompatible with the expectation to operate at a level without the benefit of strong foundations of basic, practical information handling experience.

Meanwhile, the “new knowledge workers” increasingly rely on technology not just to bestow them with the facts and information they need but also to skilfully manipulate it into a finished product.

Does it matter?

What happens when the technology fails?  Who has the knowhow or experience to check the product is accurate and is as expected? What happens if it fails the quality check?  Who figures out what went wrong?

Technology is a wonderful thing; I really do love many new technologies.  Organisations are recognising the value of people and particularly those with tenure and the depth of understanding they bring to the business; but we cannot be complacent.  When the technology fails, there is growing dissatisfaction with the lacklustre quality of services; when a problem arises, it requires depth of knowledge and experience to fill the gap.

A number of professional services organisation have begun re-aligning their KM work with Talent Development.  Recognising that knowledge and knowhow are part of the intellectual capital of the organisation.  Acknowledging that experiential learning associated with employment is something to nurture and pass from person to person, not programmed into a machine and regurgitated ad infinitum.  This is especially the case when these standardised routines appear at odds with the need to differentiate an offer by building bespoke solutions to meet specific needs and expectations.

I remain optimistic that our industry will respond and reposition in light of continuing advances.  Unfortunately, this is only one part of the equation.  If we are to thrive, we must continually demonstrate our value to convince our leaders that we have a place in the future of our respective organisations.

First issue of 2018 now out online

Our March issue contains a number of papers with the general theme of looking at the effects of technology on information and knowledge management. Hal Kirkwood returns to look at how artificial intelligence (AI) is affecting information professionals and their job roles. Delphine Phillips and Mark West from Integreon look at the future of Business Information Services (BIS) within the financial services sector and the effects of technology and other internal and external environmental factors in that area. We also see a contribution from Gabriela Labres Mallmann, a PhD student at the School of Management, UFRGS, considering the influences of Shadow IT on knowledge sharing. Here is a short overview of each of the papers in this issue.

  • The Current State of Artificial Intelligence and the Information Profession: Or Do Librarian Droids Dream of Electric Books? – Prof Hal P Kirkwood, Purdue University.

Hal begins by observing that while there has been an increasing interest in AI in the last 12 months, there has been 100% increase in the use of the terms AI and librarians. AI as a technology is fast moving from science fiction to reality with the rising popularity of voice-activated tools such as Siri to the developing use of self-driving cars and even a self-operating grocery store! His article, unlike others, is not a review of the good and bad sides of using AI, but about considering how the technology is developed and its psychological impacts. A lot goes into the development of the technology, it is not created as ‘all knowing’. It requires a lot of human interaction and consideration to develop the algorithms, providing ‘good’ and ‘relevant’ information and data to the AI tool in order for it to provide an effective service. It still also requires ‘policing’ to ensure that information it provides is accurate and relevant which still requires human interaction. His article also reviews what is being done around the world to consider the impact of AI and ensuring that it is used for the greater good rather than creating a negative impact on people and society at large.

  • Exploring the Future of Business Information Services in the Financial Sector – Delphine Phillips, Knowledge Solutions Manager, Integreon, and Mark West, Operations Director, Knowledge and BIS, Integreon.

Delphine and Mark have conducted a highly interesting research study on the role of BIS within financial services and its future in light of changing internal and external environmental factors. Their research is gathered from global investment banks and equity houses and considers the role technology is playing in the development of the BIS of the future. They review different operating models, how these are affected by internal and external changes and look at future drivers and future scope developments. They also consider the influence of knowledge management services on BIS, how they link and interact.

  • The Influence of Shadow IT Usage on Knowledge Sharing: An Exploratory Study with IT Users – Gabriela Labres Mallmann, PhD student at the School of Management, UFRGS.

Gabriela presents a new look at knowledge sharing from the point of view of ‘Shadow IT’ (software and hardware not authorized by IT departments) and its effects on knowledge sharing. The research is gathered from a series of interviews with IT users looking at how they share knowledge and information, why they share it in this way and considerations for managing risk for the future.

  • Knowledge Management Process Arrangements and Their Impact on Innovation – Eduardo Kunzel Teixeira and Mirian Oliveira of PUCRS, Rio Grande do Sul, Brazil, and Carla Maria Marques Curado of ISEG-UL, Lisboa, Portugal.

Moving away from technology and focusing more on process, this paper discusses the impact knowledge management process (KMP) has on facilitating innovation. The authors look at how different processes and different combinations of processes can affect innovation. Their conclusions, overviews in the abstract, provide a good taster of the paper itself –

1) it was identified that in general the companies apply balanced KMP arrangements;

2) that the same innovation results can be achieved using different KMP compositions; and

3) that KMP investments tend to reach a maximum effect, beyond which innovation decelerates.

  • Out of the Box – Virtual Realities in the Business World

Luke Tredinnick reviews the emergence and current uses of virtual reality technology and considers how it can impact our world. Will it become just another passing fad like 3D television or is it set to be one of the next disruptive technologies on the horizon?

  • Perspectives

Martin White returns with a review of the latest papers across Sage which could be of interest to you. Highlighted is a paper on the importance of being allowed to make mistakes in order to develop knowledge and innovate. Martin draws from his own background to illustrate the importance of this in the work environment.

Other subjects covered include the use of language and the ability to analyse and use it to consider cultural fit within an organization; considerations for HR and prepping the workplace as the amount of knowledge-led work increases with the working environment becoming more and more complex; AI and human interaction and the development of shared mental models to facilitate future developments; a discussion on the impact of libraries’ ISO standard; and the importance of user interfaces and display of search results in a meaningful way to improve findability. Luke Tredinnick and Claire Laybats

See more online here http://journals.sagepub.com/doi/full/10.1177/0266382118762967

 

Automation and AI – What does the future of work look like?

Author: Steve Dale BIR Editorial Board Member

Our news and activity streams are buzzing with articles, blogs, analyst reports and social media hype around the topic of “AI”. It’s a fairly loosely defined topic that covers an enormous spectrum of disciplines, from big data and predictive analytics, to machine learning, natural language processing, automation and robotics. Depending on who you listen to, it’s either the most important technological breakthrough since the invention of electricity, or it heralds the end of civilisation as we know it! Extreme scenarios are most certainly fantasies and should be discounted. The most likely outcome is neither extremely negative nor extremely positive.

What tends to focus our attention are the stories about how AI and “intelligent” machines are replacing roles, jobs, or even professions. What is the real truth behind these stories?

There is no doubt that workplace automation is becoming more widespread, and today’s AI-enabled, information-rich tools are increasingly able to handle jobs that in the past have been exclusively done by people (including tax returns, language translations, accounting, even some types of surgery) – automation is destined to have profound implications for the future world of work.

McKinsey recently reported that 30 percent of activities for 60 percent of occupations are now technically automatable.

Recent advances in robotics, machine learning, and AI are pushing the frontier of what machines are capable of doing in all facets of business and the economy. Physical robots have been around for a long time in manufacturing, but more capable, more flexible, safer, and less expensive robots are now engaging in ever expanding activities and combining mechanization with cognitive and learning capabilities—and improving over time as they are trained by their human co-workers on the shop floor, or increasingly learn by themselves.

Massive amounts of data that can be used to train machine learning models are being generated, for example through daily creation of billions of images, online click streams, voice and video, mobile locations, and sensors embedded in the Internet of Things. The combination of these breakthroughs has led to spectacular demonstrations like DeepMind’s AlphaGo, which defeated a human champion of the complex board game ‘Go’ in March 2016.

New milestones are being achieved in numerous areas, often with performance beyond human capabilities. In 2016, for example, Google’s DeepMind and the University of Oxford applied deep learning to a huge data set of BBC programs to create a lip-reading system that is more accurate than a professional lip-reader.

There are numerous examples of how machine learning is being used to augment human decision making in healthcare, aircraft maintenance, oil and gas operations, recruitment, insurance claims processing and law. There is barely a sector that is not engaged in some way in exploring the use of AI and automation technologies to improve productivity or accuracy.

One of the more practical roles for AI over the past few years has been to automate administrative tasks and decisions. Companies typically have thousands of such tasks and decisions to perform, and it was realized that if they could be expressed in a formal logic, they could be automated. A key feature of this type of automation is machine/deep learning and robotic process automation (RPA) – which, contrary to its name does not involve actual robots; it makes use of workflow and business rules technology to perform digital tasks.  The technology makes it relatively easy to automate structured digital tasks that involve interaction with multiple information systems.

So, what does all of this new technology mean in terms of jobs? Most analysts are agreed that whilst many routine tasks and functions – both physical and cognitive – are being automated, this does not necessarily mean that we are heading for mass unemployment as the machines take over. Perhaps one of the most extensive research programmes into the impact of AI on jobs and skills has been undertaken by Nesta. It has published its findings in the report:  The Future of Skills: Employment in 2030. Well worth a read. The report highlights that:

  • skills that are likely to be in greater demand in the future include interpersonal skills, higher-order cognitive skills, and systems skills.
  • the future workforce will need broad-based knowledge in addition to the more specialised skills that are needed for specific occupations.
  • dialogues that consider automation alone are dangerous and misleading since they rarely take account of globalization, an ageing population and the rise of the green economy.

Perhaps the last word on where AI and automation is having (or will have) the most impact should go to Gil Press at Forbes, who identifies the sectors and functions as follows:

  1. Customer Self-Service: Customer-facing physical solutions such as kiosks, interactive digital signage, and self-checkout. Improved by recent innovations such as better touchscreens, faster processors, improved connectivity and sensors. A prime example is the experimental Amazon Go convenience store.
  1. AI-Assisted Robotic Process Automation: Automating organizational workflows and processes using software bots.
  1. Industrial Robots: Physical robots that execute tasks in manufacturing, agriculture, construction, and similar verticals with heavy, industrial-scale workloads. The Internet of Things, improved software and algorithms, data analytics, and advanced electronics have contributed to a wider array of form factors, ability to perform in semi- and unstructured environments, and the “intelligence” to learn and operate autonomously.
  1. Retail and Warehouse Robots: Physical robots with autonomous movement capabilities used in retailing and/or warehousing. Amazon deploys this technology throughout its warehouses.
  1. Virtual Assistants: Personal digital concierges that know users and their data and are discerning enough to interpret their needs and make decisions on their behalf.
  1. Sensory AI: Improving computers ability to identify, “understand,” and even express human sensory faculties and emotions via image and video analysis, facial recognition, speech analytics, and/or text analytics.

He goes on to say:  “There is no question that we will continue to see in the future the same disruption in the job market that we have witnessed in the last sixty-plus years of computer technology creating and destroying jobs (like other technologies that preceded it). The type of disruption that has created Facebook and Tesla. Facebook had a handful of employees in 2004 and today employs 20,000.  Tesla was founded in 2003 and today has 33,000 employees. Whether AI technologies progress fast or slow and whether AI will continue to excel only at narrow tasks or succeed in performing multi-dimensional activities, entrepreneurs like Zuckerberg and Musk…will seize new business opportunities to both destroy and create jobs. Humans, unlike bots and robots (now and possibly forever), adapt to changing circumstances.”

One thing we can be sure of: the rate of change will continue to accelerate, and if we wish to remain relevant in our chosen professions, we need to identify and refine the skills that can’t easily be automated. Whether that’s a shrinking or expanding environment remains to be seen.

Man vs. Machine vs. Data….?

Author: Penny Leach SLA LMD Past Chair 2017 and BIR Editorial Board Member

Please note this post contains the personal views of the author and are not connected with her employer
 

I recently had the pleasure of attending the 14th Perfect Information Conference (PIC2017) in Bath, England.  This annual event, hosted by the company Perfect Information (part of Mergermarket), brings together leaders and senior members of information services from within financial and professional service organisations with representatives of their content and service vendor partners.   The high number of repeat attendees confirms the conference’s value.  This year’s programme theme was ‘Man vs Machine: comrade or threat’.  For me (spoiler alert!) the whole event reaffirmed the current and future value and potential of humans in an increasingly technological world.

The conference programme includes keynote speakers, more practical workshops and hot topic think tanks (and of course some socialising!).  What seemed to me initially a rather disparate set of topics actually transitioned from the big picture of artificial intelligence (AI) and its future to more practical implications of change for businesses today.  Having worked myself for a short time at the (original) Turing Institute in the early days of AI, it was fascinating to hear where AI is today.

AI is all around us, was the clear answer from the three speakers who focused on this topic, respectively Marc Vollenweider (Evalueserve), Anton Fishman (Fishman & Partners) and Nicolas Bombourg (Report Linker).  Marc, who is transitioning from CEO to Chief Strategist of Evalueserve, spoke about the explosion of data sets, and the business value to be derived from cheap but effective analytic use cases.   Anton alluded to the ‘perfect storm’ of converging technologies that is affecting the world of machine learning.  Nicolas described Artificial Narrow Intelligence (ANI) – where we are now (machines specialising in one are) – and how we are moving closer to Artificial General Intelligence (AGI) – machines thinking like humans – and even beyond to Artificial Super Intelligence (ASI).

Are we heading for dystopia or utopia?  There were references to sobering statistics about the predicted negative impact on job numbers, for example, Mark Carney’s speech on the  ‘hollowing out of the middle classes’ and Frey & Osborne’s research in to the future of work in the US.  Ultimately however the message was upbeat. Marc is definite that ‘insights need humans’, and has written about the benefits of combining mind+machine.  Anton referred to the opportunity for the ‘rise of humans’ that Microsoft’s Envisioning Officer has described.   The message is that technology is supporting humans, expanding our potential – AI is already invisibly enhancing our world.  This is not a zero sum game for mankind, even if it does create much management uncertainty, ethical dilemmas,  job redefinition and the need for a new ‘social contract’.

What were my key takeaways from these speakers and all the other interaction at the conference for me in my role as an information professional and services manager?

Information professionals do have roles to play in the new data economy, where the flow of data is driving innovation and growth, as long as they are open minded and upskill.  Marc has elsewhere talked about the emerging role of the information engineer in creating analytics solutions. McKinsey recently re-recognised the need for translators between technology and information, reaffirming the need to link IT, understanding of data, and business need.  Establishing the veracity of data is of course a traditional information professional skill.

Information professionals need to engage with the business, via new channels such as their workplace’s Chief Information or Data Officer (CIO, CDO) – wherever analytics are happening – and change the scope of their services to help the business build effective productivity tools and  new trigger-based workflows, and avoid data lakes that become data graveyards.

Change management is important – keeping people engaged, attracting new talent, enabling career progression, as well as ensuring effective use of the new tools and data by the business.

And for those directly engaged in buying and selling data there are reminders of the early days of the internet and outsourcing in the challenges of delivering and consuming data in new ways – for the vendors, what to build first for which client, how to protect the data, how to charge for it; for their clients where to focus efforts, who will eat the costs; and for both parties, how to deal with the increased visibility of data quality issues.

Overall the Conference ended on an optimistic note in contrast to the anxieties of the 2016 Conference (as described by in the opening session of the conference by Robin Neidorf) and inspite of the seismic political changes we have seen in the UK and USA in the last twelve months.  It will be interesting to hear how things have further changed for the attendees by the same time next year.

Let me know what you think of AI’s impact on your world.

Penny Leach

SLA LMD Past Chair 2017