Artificial Intelligence and Our 5 Senses

Artificial Intelligence (AI) is very much the thing of the moment. A lot of new enterprise models appear to be built on a combination of AI, IoT and Big Data to collect and process massive amounts of data and to apply human like skills of recognition and decision making with a level of consistency, speed and volume that was previously unattainable with conventional processes or systems. This leads to greater insight, improved performance, better use of scarce skills and consequently other benefits such as improved customer experience, better security and higher cost efficiency. It also means that some tasks which would never have been practicable before are readily delivered via automation.

However when we look at what a human can do, AI capabilities are still lacking. The human capability to use our 5 senses, think abstractly, innovate and test hypotheses, to develop new learning are still a challenge to AI.

Just focusing on the 5 senses: Sight, Hearing, Touch, Smell and Temperature, AI's progress is quite patchy.

In the past 30 years or so, AI's ability to deal with vision and speech has progressed from a miserable score of  about 1 to 2 out of 5, to a 3 to 3 and a half out of 5. Handwriting recognition works well with typed characters, but still has a significant error rate with hand writing. Face recognition systems work quite well with caucasian faces but apparently struggle to deal with black african faces. It is not clear to me  whether the latter is to do with poor training (i.e. not been given enough sample data for adequate deep learning), fundamental flaws in the assumptions made when designing the systems to recognise features or some problems with technical measurement of contrasts and camera sensitivity across the visual spectrum. Whatever the problems, error rates are still quite high.

Temperature is quite easy with thermo-couples and can be engineered to deal with temperatures that people could not with stand. So high marks of 5 out of 5 are to be expected.

Touch has been a major area of research ever since robotics became a hot topic in the late seventies. Most of the work so far has focused on sensing around grip, so that a robot could pick up an egg or some other delicate object without crushing it. But I have not seen anywhere the type of sensing that could differentiate between the touch of skin, rubber, silk, cotton and wool. So this definitely languishing in 2 out of 5 territory.

Smell has barely been touched. However some work is suggesting that techniques for smell processing, which mimic natural biological processes could be invaluable in extending AI capabilities to deal with situations where learning data is limited, ambiguous or masked with other "noise". Smell has potentially many applications in medecine, agriculture and food processing. The techniques however, may be useful for Autonomous Unmanned Vehicles which often have to deal with unforeseen situations or noisy environments.

So there still a long way to go for AI to reach science fiction like capabilities. Conquering the 5 senses represents the first step. The question is, how long will this take and will all humans be machine augmented, by the time we get there?

Aligning Digital Exploitation with Operational Capability

All Enterprises Are Not the Same

One of the tenets of Digital As Usual (DAU) is that most businesses are not pure play digital organisations. In almost all cases there are operations and activities which are essential to either fulfilling service or delivering a product which need to be managed as well. The Digital Adoption Framework assumes an Industry Categorisation Model as illustrated below.

The Industry Categorisation Model identifies 5 Categories along a continuum of Industry Product Nature. At one end of the continuum are “pure play” digital organisations, whose products are Information and Content based. These consist of Market Platforms like eBay and Compare the Markets.com who provide a platform, through which other organisations and people sell their goods and services, as well as Publishers and Providers of Content who provide videos, e-books, podcasts, news, music etc. At the other end are Extractor and Grower type businesses, such as coal mines and cocoa growers who provide natural unprocessed products. Off course not all businesses fall neatly into just one category. A major energy company such as Shell or Exon may straddle several categories along an Upstream (exploration, drilling and extraction) to Downstream (transport, refining and distribution) activities.

So why is this important when thinking about digital businesses? Well each different type of business operates in different ways, has a different level of dependency on capital assets and has different opportunities for exploiting digital technology to augment its business model. Additionally, the whole concept of Lean Delivery via Design Thinking, Agile Delivery and DevOps may have different opportunities and constraints as well as nature of delivery in each category.

Some Examples of Exploitation

The diagram below looks at the potential for exploiting Artificial Intelligence (AI) technologies according to Industry Category.

The “X”s in the table represent significant opportunity for exploitation against a generic type of Use Case. So for example AI technologies (machine learning, visual recognition, natural language etc.) can be used to provide insight and learning in every type of organisation, but there is very little opportunity for them to add to “customer experience” in say mining or farming. Again there is huge potential to build AI capabilities into consumer products, e.g. Alexa in Amazon’s Echo, but there is little opportunity to enhance a product such as iron ore or tree trunks produced from a Natural Product Extractor or a Grower. 

Alternatively for IoT, see diagram below, the opportunities are pretty limited for an Information publisher, but there are many areas where a farmer could use it to track livestock and monitor soil conditions and crop conditions, as well as control assets such as vehicles and trailer equipment.

This moves us onto exploiting iterative approaches within an organisation. The other night I was talking to someone from a major pharmaceutical company. Drugs take years of research to identify promising candidates for a problem and to test for efficacy and safety, as well as meet regulatory demands. There is little opportunity to exploit Iteration in this type of product development. But for the company overall, there are opportunities around internal processes and with Agile Marketing.

Agile Marketing is a growing area of adoption within industry and works well for consumer driven businesses. At its core is the use of small integrated product teams which focus on market research, product promotion, delivery of marketing and promotional materials, advertising and the information analysis systems used to support marketing data analysis. These teams tend to work in multiple sprints testing out marketing promotion hypotheses and rapidly identifying which approaches produce optimal results, to improve overall revenues. 

Iterative development also its quite different for delivering an oil tanker or a film, to the development used to support a consumer service. In the case of the former, iterative design, development and continuous testing techniques are used to deliver a single integrated complex product over a period spanning many months if not years. In the latter they are used not just to deliver the product, but to continuously keep it fresh, up to date and moving ahead of competition that is playing catch up.

So it's important not to adopt a one sized fits all approach, but to really get down to what can work for your organisation. This is part of what I talk about in my book.

The Trivia of IoT

You know that something has become too trendy when you key in the subject and some other random item and find that you get lots of returns. So yesterday, I tried it out with IoT.

So what do you get with IoT and Bathroom? Well a polemic on the bathroom of the future. Yes, we will have smart toilets which are not only eco friendly but diagnose medical conditions: Pregnancy, bacterial conditions, PSA levels, sugar levels etc. Smart showers are already here for the eco-friendly trendy set. They minimise wasted water, go straight to the temperature that you like and give feed back via LED lights on how much water you have consumed. Then there are smart toothbrushes which give you feedback on toothbrushing technique. Finally,  there is the smart mirror. This allows you to catch up on the news and weather forecast, whilst brushing your teeth. You could also monitor any of your smart household devices, such as your smart toaster, at the same time whilst also checking incoming messages. Sounds like hell doesn't it?

So I decided IoT and Underwear would then be a suitably trivial query. Well you get responses on the usual fitness app equipped wearables for monitoring pulse and other vitals whilst exercising. There's also smart yoga pants, which provide feedback to help you form the right posture when exercising. But there is also a quite exciting smart bra device, the iTbra, which can help monitor the wearer for signs of cancer. This has the potential to replace the expensive, stressful and quite uncomfortable process of having mamograms with something which monitors circadian temperature changes to detect abnormal patterns. Apparently this is based on a proven technique which has existed for decades, but was previously too cumbersome to use, due to the need for lots of cables going to instrumentation. IoT does away with the cables and allows for a potentially more accurate test which can be completed within a few hours of monitoring.

So, I then started thinking like a teenager and yes there is a smart tampon. The my.Flow  is there to help avoid the embarassement of blood staining the wearer's clothing, which would be a mortifying experience for anyone. As a man, I don't think that I should comment any further. But moving on to men, yes you can indulge in the smart condom, or i.Con. Actually its not a condom, but a ring you can slip over the top to monitor your performance statisitics during the act of copulation. Do we need it? is it useful? I'll leave it to you to decide.

So what do we learn from this trivia and is there a serious point. Well, its obvious that soon we will not be able to escape the World of Interconnected Smart Things (WIST). It has become pervasive across almost any field that you can think of (including farm fields of course). With it comes the danger of people who want to exploit it perniciously. So we need to start thinking about protecting ourselves and privacy and security in our homes and daily lives. However, this goes beyond the simple steps of securing your WiFi, protecting your passwords and identity, to seriously ensuring that your networked home and persona are well and truly protected in a systematic way and that you have separation of "concerns" and firewalls between your virtual world and the rest of the world.

The Return of the Crackberry

This week, Blackberry held its annual security conference in London. so it was a good time to catch up with what Blackberry is doing now, after the melt down of its original secure corporate phone and email business model. 

Well, it is still in the phone business. Partnering with TCL, who undertakes hardware manufacture and smart phone distribution, blackberry is still designing new smart phones. The new models look very smart and offer a key pad equipped, touch sensitive android phone at a competitive price point, i.e. several hundred notes cheaper than the equivalent Samsung or Apple model. It would be easy to dismiss them as being a choice for someone who is into retro chic or an old school aficionado of key pads. However, there are some subtleties in the product design which are increasingly relevant in the current environment of aggressive cyber crime. The devices are designed and built bottom up to be resilient to infiltration and takeover, combining both hardware and software features for this. Additionally, the thumbprint security control is implemented in a way which allows the user to differentiate between what is private and what is shared. Which coupled with features in Android enabling separation between personal and corporate personas, makes it very much a smart phone of the age.

However, mobile phones are really only a side show and not the main story. Blackberry has built on its global secure telecommunications capabilities to emerge as a leading "Enterprise of Things" (EoT) enabler. Someone described its primary mission as being to "Secure your Communications and your Data". However the vision is one of comprehensively enabling secure IoT exploitation by enterprises, leveraging industry standard to deliver on government quality security in the deployment of Smart Things, or as one of the speakers stated "Moving from Mobile First to Things First".

There's a comprehensive set of products with SDKs which enable OEMs to develop secure smart products and Enterprises to deploy Thing based business models securely, whilst users enjoy a relatively seamless secure access experience across a comprehensive range of products.

Although what was probably the most impressive aspect of the conference was to see representatives from Google, Microsoft, Samsung and Blackberry discuss how they are collaborating to deliver a new generation of products which integrate securely and build the IoT world of the future.

Digital Era Customer Contact and Experience

In the 1980s, The theory of Customer Care was propogated widely at the same time that call centres started to evolve into contact centres. The subtle difference between a call centre and a contact centre being that a contact centre is supposed to deal with customers through all media of communication: letters, telephone calls, and direct contact in person if somone knocked on your door. With the '90s, email was added and then websites and latterly mobile apps have started to grow traction.

In parallel with this we have seen all sorts of mechanised and automated technologies arrise to help improve productivity and efficiency. IVR or Interactive Voice Response (which really is not that interactive), Power diallers to fuel direct telemarketing campaigns, backed up with scripting to ensure consistency of message were also introduced followed by features such as distributd call centres, off shoring, workflow based load balancing etc.

The touble with all these approaches has often been that they focus on work load balancing, limited levels of scalability and cost efficiency rather than personal experience. Offshoring also intorduced problems with culture and language, as well as greater possibilities of fraud and exposure of customer data to theft and exploitation, as "solid legal contracts" have proven to be paper tigers in protecting confidentiality. In fact in some countries, such as Singapore, which have reutations for innovation and fast moving companies, customer experience is terrible.

This just does not cut it in the modern digital age, when customer experience and reputation will make or break a company. We now have people pondering the future of the contact center and whether there will be any people in them to take calls with the advent of AI and chatbots as well as the use of IoT to collect customer behaviour information and change the model by which companies interact with customers.

The real issue is that many of these advances can be uselfull exploited in situations which suit "self serve" activities such as "how can I fix my thermostat or change my account settings". They don't particularly work well when there has been a service failure or a billing mix up. Assumptions that Generation Next Letter in the Alphabet will always want to contact you via a certain technical chanel are also wrong. Peoples contact method decisions are often affected by context. If they are not allowed to use mobiles at work for instance, or they are in an airport and have lost their mobile phone or iPad and urgently need help with a travel problem, then other contact options are needed.

The new technologies offer opportunities, but greater effort needs to be put into creatively exploiting them and designing a great customer experience. So it is interesting to see that a company like Creston is establsihing a global network of Customer Experience Centres to show exactly what it can do for people. It will also be interesting to see how Creston exploits this interaction chanel to improve its customer offerings as well.

Is AI a threat to the Rockstar Developer?

This week IoT TECHEXPO was the main event at Olympia in London, combined with BLOCKCHAIN EXPO and AI EXPO. This was an impressively large event, though still not quite at the critical mass of some of the Security Exhibitions held recently at the same location. Though it shows that converged application of the 3 different technologies is gaining traction.

It was impressive how many vendors were launching new services based on blockchain. An aweful lot of them were based around assuring identity and provenance of assets, people and devices. So there were universal "passport" service, property dealing services, fine art trading, new travel disintermediation services and so on. In addition to several crypto currencies that I had not heard of before, there were a lot of wallet and token services.

In fact at one point I got the impression that the suppliers there were providing the means for circumventing national authorities with their own private currencies and virtual passports. For if you imagine a consortium of multinational businesses using tokens to trade with some end of year settlement in real money, this could totally confuse any picture of their tax obligations based on monetary trading transactions.

Putting this aside, the most interesting talk was probably the one on using AI for programming by Kiki de Bruijn of Diffblue which was heavily oversubscribed. She described the challenges of trying to generate code without human input and advocated the use of AI and Machine Learning based tools as productivity and quality aids instead, as it is difficult to replace the creative aspect of development as well as deal with all the potential complexities of developing well integrated code to meet all eventualities,

Moscow Goes IoT whilst Businesses Struggle with Digital Culture

The City of Moscow recently announced that it had started to manage the operations of its whole fleet of 22,000 vehicles using a mixture of IoT and Machine Learning to optimise deployments for both effect and efficiency. This covers everthing from refuse trucks to snow plows. Efficincy savings of something live 9M roubles per month are now being claimed or roughtly £150,000. At the same time it meant that Moscos coped admirably with the recent extreme snow conditions, keeping the city moving with knock on gains to the lives and economic benefits of the local community. This is a good case study for Smart Cities.

At the same time, the digital work is throwing up whole new social and quasi religous groups such as the Kult of Kek and Witchblr. These groups have their own approaches to changing a frustrating world, but represent powerful pressure groups whose influence on things like social choice and consumer behaviour is leveraged well beyond their actual membership. They are however, non conformist and totally confusing to the pre-millenial generation, fusing a whole set of religious, game and folk themes with strong political messages and a dash of geek humour.

Little wonder then that a recent survey (by Virtusa and Forrester) show that 85% of businesses have digital plans, but many don't really know what to do next. They have not really grasped the principles or developped the capabilities to pursue their agenda. So they are not yet prepared for the fact that Digital makes access and on boarding of customers easier, but businesses can no longer rely on 75% of business being from repeat customers, unless as I pointed out in The Way of DAU they adapt to relentless renewal and re-invention and manage their ethical stance to keep their brand shiny and clean.

Has Big Data Lost its Mojo?

A number of surveys were published during 2017 which suggested that Big Data has lost the CxO mind share that it had.

The overall impression gained was that business leaders are putting their emphasis on Artificial Intelligence (in particular Machine Learning) and IoT, whilst CIOs and Technical Leaders are worrying about Security and Lean (Agile and DevOps). 

At the same time the consumer and gadget end of things are focusing on wearables, voice and VR/AR based devices (implying other types of AI are getting important). 

Somewhere in the mix, people are worrying about culture, product management and marketing, and organisations designed around empowered product teams.

There almost appears to be an assumption that big data has been cracked and apps are easy. Also, data governance is not really getting the attention that it should do and vendors are pushing APIs for integration.

This suggests not just a gap between business vision and technical capability to deliver, but also that the consumer led boom in exploitation is going places which don't necessarily fit well with classical business environments. Open plan offices are not the place where people should all be talking to their digital assistants.

The lessons I draw from this are than business management teams need to start thinking holistically about what is needed to deliver their new business (i.e. digitally augmented) strategies and models, but also about what the future workplace looks like. Is the office finally dying?

They also need to get real about data. It needs to be managed using lean data principles. Integration and security are vital to get right. IoT and other exploitation approaches are going to emphasise Big Data's importance further. So it's vitally important to develop data specialists who understand your business and are networked with the right people. A recent article in the Harvard Business Review suggests that most organisations have focused on technicalities and need to look at a more immersive approach and organisational issues too.

IoT Adoption Drag

IoT Makes Slow Progress

Despite all the hype and general acceptance of its possibilities, IoT is making slow inroads into enterprises and can still be regarded as being in the early period of adoption in many businesses.

I have talked about CxO behaviour and culture in previous blogs as well as security and other issues, so what is holding IoT back?

Business Leader Perspective on IoT Adoption

A recent survey by the Economist Intelligence Unit (sponsored by IBM and ARM) involving over 800 executives around the world looked into the issue. This suggested that fewer than 10% had extensive IoT implementations, although many had dipped their toes in the water.

More positively, around 25% stated that IoT adoption was sparking innovation within their organisations with 22% indicating that IoT offered new opportunities, 20% saying it was changing business models and 16% believing that it was enabling entry into new markets and initiatives. So the overall picture is quite positive about its transformational potential.

Only 15% believed that IoT would significantly reduce costs, suggesting that IoT adoption is more about re-imagining their businesses than old fashioned cost cutting.

Interestingly enough, few executives saw technical issues as blockers to adoption. The key constraints cited were the Size of Investment needed (29%), Security and Privacy concerns  (26%) and lack of CxO Knowledge and Commitment (23%). 

This points to a smoking gun around Cost and Executive Behaviour / Team Buy In.

IT Leader Perspectives

Interestingly enough the British Computer Society's (BCS's) survey on Digital Leadership shows some mismatches.

Over the last 4 years the principle concern of IT Leaders in the Digital Space has been Security (averaging 60% of responses, plus or minus 2% over this period). The second most important concern has been Cloud (at 50%). In the last 2 years, Governance has moved into third place at 35% followed by Agile and Big Data at around 25-30% each. IoT was only cited by respondents as a major interest by 18%. Although Cloud, Agile and to some extent Big Data are involved in IoT adoption.

This suggests 2 things. Business Executives are over confident about implementation issues and that IT Leaders have a mismatch in alignment and understanding with business aspirations.

IoT Success Factors

So it appears that CxOs need to get together and agree a common vision and set of priorities. IT Leaders need to build IoT into their plans and there is a need for education and collaboration to get a coherent approach in place.

The Value of Data

Today I realised that the evidence for Digital being the New Normal was compelling when I read about the way that the agricultural industry is embracing IoT, Big Data and Machine Learning (see: FoodIT: Fork to Farm ) which provides comprehensive overview of the 4th annual conference examining how to join the ecosystem from consumer to farmer up, so that farms respond to consumer tastes, improve yields, avoid waste and provide the provenance etc. that consumers now want. There are great opportunities to provide the next Deliveroo for farm products and by-pass the super markets entirely. This would greatly improve profitability for farmers and the experience for customers as they get fresh, seasonal, sustainable, quality products directly. It could also bypass the scandal of the modern monopoly of slaughter houses, if the right investments were made.

I especially enjoyed the redefining IoT by the conference as the Internet of Tomatoes, spelling out how much the industry has got into it.

There were also some interesting predictions for Data in another couple of articles on the data makes possible  site by Kirk Borne of Booz Allen Hamilton about trends for this year. Personally I found his comments on graph analytics the most compelling as this is an area of growing maturity in which it is now possible to easily uncover insights into linkages of cause and affect and networks of people and activity. He also had interesting things to say about hyper personalisation and realigning AI with a people centric model of mutual assistance to do more (similar to some of my previous observations). Finally he coined a new phrase DataOps to describe lean development approaches as applied to big data.

Digital Adoption Framework

A lot get's talked about Digital, but there are few comprehensive approaches to adoption available for reference. 

This is why I was interested when I came across Vadrim's DAF diagram, reproduced below. Enjoy!