Wednesday, October 31, 2018

Machine Learning, AI and Digital Transformation

AI refers to a variety of technologies, some pretty basic and others very advanced. Basic AI has been in use for many years. As for the advanced - well, it's hard to define that, since it keeps changing so fast.

A powerful use of AI now is in the area of autonomous data analytics. The word autonomous implies that the analytics are able to stand on their own - able to make their own decisions. This is what is happening with the help of AI and machine learning.

Companies have more data available to them than at any time in history. However, that data is of little use unless it is analyzed so as to yield useful insights and prospective information. Also, the analysis must go a lot further than simply being sufficient to support a particular hypothesis. Rather it must be mined to yield the secrets and lessons it holds.

Autonomous analytics, aided by AI, supports that approach by being able to recognize relationships in the data that can then be used to formulate further lines of enquiry. That's where machine learning comes in. All of this is automated. In addition, the analysis can encompass very large bodies of data, maybe even all the data available to a particular company.

Techniques for implementing such systems are advancing quickly and several companies have been reporting useful results.

For a good paper on this subject check out this one.

Friday, October 26, 2018

AI and the Trough of Disillusionment

Artificial Intelligence adoption is playing a major role in the implementation of digital transformation by businesses. The hype is strong and expectations are running high.

Recent Gartner surveys show that a major part of AI implementation is likely to be in the area of customer interaction. It's not going to happen in the next year or so. Five to ten years is being put forward as the most likely timeline. Can the hype survive such a wait? Not likely.

AI still needs to go through the Trough of Disillusionment as identified in Gartner's Hype cycle. With so much emphasis on customer interaction, that trough could be deep. Past and current attempts to have customers interact with computers have been rife with frustration on the part of the customers. Even the established and simple process of calling a company and being met with an array of choices in the form of selecting numbers still does anything but encourage customer interaction.

Of course, the new AI systems will be a lot better than that and will be conversation based. But they had better be very good. Anything less will bring us back to our experience with the more primitive systems of the past. And the trough of disillusionment could then be a canyon. For more on Gartner's take, check this out.

Monday, October 22, 2018

Facial Recognition - A technology to watch

There's a growing realization that facial recognition (FR) technologies, aided and abetted by AI, are becoming more powerful, will be used more often and pose a significant threat to privacy.
In the past, tests have shown that facial recognition is often not very effective. For example, at Boston Logan Airport, volunteers posing as terrorists could only be successfully identified 60% of the time during a test, a rate that was determined to be unacceptable.

The AI component is becoming more powerful, however, and performance will be better in the future.

Given the widespread interest in using FR for a variety of purposes, a major purpose being safety, its use will increase rapidly. Along with this will be a need for new regulations to protect individual privacy.

It will not be possible to seek permission for all applications, so other means will be required, such as oversight boards, and notices to inform people that FR is in use. It will be necessary to define when permission will be required. Or when people may be allowed to wear sunglasses. Lots of issues will arise. See this article for example,

The use of AI will likely be useful in dealing with privacy, by introducing logic that deactivates FR in certain circumstances.

The privacy implications of FR is a major issue in the battle of technology vs privacy and developments merit close scrutiny and monitoring.


Saturday, October 20, 2018

Some Outstanding IT Trends for 2019

Gartner recently released their much anticipated list of IT strategic predictions for 2019. There are a few trends that stand out.

In 2019, efforts will continue to adopt AI, but this will be a rocky road because of skills shortages. Although it may begin to improve by year's end. Also, AI will likely enter into emergency care of chronic patients through powerful AI screening techniques. Thus freeing up emergency rooms for true emergencies.

Privacy will continue to be under attack on several fronts. Increased use of facial recognition technologies will mean that roadside cameras and mobile phones will be able to identify people easily. Think about that. However, Gartner anticipates that public security monitoring will increase without opposition because of fears about public shootings. Also, privacy in social media will continue to be a problem which will remain unabated because people will continue to use social media.

Also in the privacy domain, blockchain has serious loopholes (particularly because it contains text fields that may not be encrypted), which will lead to more privacy issues as blockchain gains more adoptions.

There are other interesting insights or 2019 and beyond. Read the full list here.

Friday, October 12, 2018

How SYSCO is moving to the Cloud

Sysco is one of the world's largest corporations involved with food distribution. In this time sensitive industry, IT systems that are reliable and timely with good data and excellent responsiveness are critical.

There is no tolerance for downtime. Cloud systems offer up parallel systems so if a system goes down, action shifts seamlessly to the other systems. There is lots of scalability so the company doesn't get caught in lack of resource.

For these and other reasons, Sysco decided to move their systems to Amazon Web Services (AWS), one of the world's foremost cloud services. Sysco was in the midst of large ERP installations and upgrades, which are major, very expensive projects with the potential for downtime, particularly when the new services go live.

Sysco opted for a "strangler" technique of implementation, which involves identifying particular sets of functionality in the legacy systems and their ERP, rewriting them as modules so as to work in the new chosen systems and then implementing them on AWS. Eventually, the old systems can be shut down.

In recent years, they have been employing this method with considerable success. Costs are reasonable, and functionality is preserved and improved.

For a summary of this experience, check out this website.

Tuesday, October 09, 2018

Advanced Technologies Need Not Scare You

A growing body of research and literature is pointing out that many traditional jobs will be eliminated or substantially reduced because of the growth in availability of big data, and use of analytics, Artificial Intelligence and Big Data. As a result, many people are concerned about their jobs.

Some will indeed be eliminated, for sure, but in most cases, they will have little to worry about. In fact, they will have some new and interesting challenges.A recent study by the World Economic Forum directly addressed this issue. It has some very positive messages.

There is nothing new about technology incursions into the workplace.We've all experienced it. Taking accountants as an example, the more routine aspects of the job - bookkeeping and preparation of financial statements have been automated for years. The more judgemental parts of the job, such as valuation of assets, have remained largely with humans. However with the use of AI, more of these judgements will be supported and/or initiated by the AI software.

Already, AI is being used to support decisions. But it will move well beyond that. The way it is likely to play out is to imagine how decisions are often made now - through groups of people with different skill sets. Imagine when one of the members of the group is a computer, or an AI engine - one who fully participates in the discussion. Perhaps think of it as a robot. It can listen to ideas, comment on them and offer suggestions. It brings to bear past decisions in similar circumstances and the results of advanced analytics.

The people are still doing their jobs, but with more useful information at their disposal. Less time is spent on the mechanics of the analysis and more on the judgements and decisions to be made. Not everyone is going to know how to do advanced analytics, but some or all are going to need an understanding of them and how to use the results. Therefore as the technology evolves, the level of the people jobs will rise to more challenging and interesting levels.

According to the WEF study, "As companies begin to formulate business transformation and workforce strategies over the course of the 2018–2022 period, they have a genuine window
of opportunity to leverage new technologies, including automation, to enhance economic value creation through new activities, improve job quality in traditional and newly emerging occupations, and augment their employees’ skills to reach their full potential to perform new high value- added work tasks."

One of the results of these strategies we will hear more about will be augmentation strategies - strategies to augment existing jobs and processes with advanced technologies.

Sounds like an exciting new time, time to accept the challenges, not to be afraid of them.

Friday, October 05, 2018

Challenges Facing the Alignment of Business and IT Goals

The extremely fast growth in the rate of change in technology is creating new challenges for management, both the traditional business managers and IT management. For several years, there has been a recognition that the objectives of IT and the business overall need to be aligned to produce optimal results. Generally, this has been addressed by ramping up the role of the IT executives in the organization - creating CIO VP positions, having the CIO report to the CEO rather than the CFO, placing the CIO on the Board and other similar organizational and cultural steps.

All of this is good, but not all organizations have implemented these steps and even those who have are facing challenges.

A major source of the issues is simply the pace of change, which is rapid and unprecedented. The realization by business managers of the importance of digital transformation has led to growing requests to their IT people, from artificial intelligence to machine learning to the impact of the internet of things to simply automating greater swaths of the business activities of the organization.

The pressures on IT have led to resource, cost and budgeting issues, which of course sends pressure back to the business leaders.

To address these issues, both business and IT leaders need to change their outlook - and skill sets.

Business leaders (those executives who have not traditionally been part of the IT community) need to gain a greater understanding of IT management issues. This can be achieved by appointing the CIO as a VP and reporting to the CEO. This makes the CIO a peer with other senior executives with the concomitant elevation in the level of discourse between them. The organization also needs to create means by which the two groups can interact  - advisory committees, working groups, etc.

Such appointments can also lead to changes in the perspective of the IT leaders. Greater involvement in business issues through the board and other committees, will shape their perspective on the demands placed on IT. Over the long run, these changes will serve to shape the executive positions themselves as well as their educational processes, with people interested in business management gaining greater IT knowledge and IT managers enhancing their general business knowledge.

This is the trend anyway, but explicit recognition of it in corporate management will help to speed the process.


Tuesday, October 02, 2018

The Scope of Digital Transformation

Digital Transformation has been defined as "the profound transformation of business and organizational activities, processes, competencies and models to fully leverage the changes and opportunities of a mix of digital technologies and their accelerating impact across society in a strategic and prioritized way, with present and future shifts in mind." (i-scoop)

Implementation of digital transformation requires a defined strategy and all of what that implies, including definition of activities, processes, timelines and responsibilities. It also requires careful identification of technologies - those in place and those that might be adopted. It requires digitization of numerous processes in the organization, including those that have never been digitized before. True digital transformation is comprehensive.

Digitization of processes require proper process management, which includes business process reengineering and change management. Inevitably the involvement of the people carrying out these processes is required to make this work. People involvement is necessary all the way through, but one of them is in areas being changed that involve interaction with customers and other stakeholders. People who are closest to customer service should be involved, perhaps even the customers themselves.

In digital transformation as currently practised, there is an increased reliance on the latest technologies - big data retrieval and analysis, artificial intelligence, machine learning and internet of things in managing corporate assets. These are prime areas where digitization might replace some or all of what people are currently doing.

It's easy to see that digital transformation is a major area of corporate management - transformative and long term. Evolution of cultures and technologies cannot be ignored.

Future articles will delve more deeply into these issues.