The Great AI Wars

The Great AI Wars

Last week saw one of the last big players make their position in machine learning and AI clearer at Apple's WWDC event with the launch of their machine learning options. These days you're not a credible large cloud provider if you don't provide some interesting APIs around machine learning and AI with the likes of Google, IBM, Amazon, Microsoft (Azure), and Alibaba (Aliyun) to name but a few. Apple's discussion focuses on being able to embed these technologies on the device with the Apps rather than perhaps the building of the models and the execution – much less focus on pushing data into the cloud.

The war I speak of in the title however, is not some dystopian future where humanity fights for survival but rather the current war over talent that enables the use of these technologies. Insurers going through digital transformations and looking deeply at their analytics are finding they are competing with ever more unlikely companies for talent including rising InsurTech firms as observed in previous blogs. The good news is that basic machine learning capability and training is increasingly available as the democratisation of machine learning continues apace – in fact if you look at Apple's documentation this discusses the ease downloading and converting models and integrating them to Apps rather than the nuances of various training algorithms.

Machine learning isn't new to insurance with coverage in our predictive analytics reports courtesy of Nicolas Michellod and case studies. It is clear however that these tools and techniques are increasingly being embedded into solutions throughout the insurance eco-system and beyond – and they are raising customer expectations. A discussion on what this means for core systems is given in my recent report here, as well as a discussion on what this means for new front end opportunities with the rise of chat bots in our discussion on conversational systems and a broader discussion on the differences in designing intelligent systems versus programmed ones is discussed in designing the aware machine.

While AI is a battleground for the big players for insurers it is becoming an increasingly accessible source of new approaches and automation – both an opportunity to better serve customers as well as cut costs. The ease with which machine learning and AI can be embedded into simple applications now will only increase adoption and there are small things any insurer can do. Of course if you want to go much deeper, as pointed out in this Harvard Business Review article, if your company isn't good at analytics, it's not ready for AI. I disagree a little with the authors perhaps, we're in a world where anyone can do something – one can just download and convert a model and incorporate it into our systems as pitched by Apple.

For those looking to go further, the good news is there are many vendors that can help, and many partners too of all shapes and sizes. I'm happy to say the InsurTech investments in the industry are only increasing this number and the opportunities for applied AI in insurance. Further, there are many conferences discussing both analytics and the rise of AI – if you're attending or looking for them do get in touch, I or my colleagues would love to discuss.

How Insurity’s Acquisition of Valen Could Create a Virtuous Analytics Circle

How Insurity’s Acquisition of Valen Could Create a Virtuous Analytics Circle
It’s open season on insurance technology acquisitions in general, and for Insurity in particular. Today’s announcement of Insurity’s acquisition of Valen Analytics is now Insurity’s fourth acquisition in a multi-year string: Oceanwide, Tropics, and in rapid succession Systema and Valen.   The potential for crossing selling among the five customer bases is obvious.   Less obvious, but of potentially even greater value, is Insurity’s ability to invite all of its insurer and other customers to use its Enterprise Data Solutions IEV solution as the gateway to Valen’s contributory database and Valen’s InsureRight analytic platform.   Insurity now has the scale and the means to create a virtuous analytics circle: individual customers contributing a lot of data through IEV to Valens and receiving back analytic insights to feed into their pricing, underwriting, and claims operations.   Good move.

The Evolving Role of Architects

The Evolving Role of Architects

In the last couple of weeks I’ve had the great opportunity to spend time with IT architects of various sorts both inside and outside of the insurance industry. The discussions have been illuminating and offer different visions and futures both for technology that supports insurers and for the future of the architecture function in insurers.

One of the main events that allowed for this conversation was a round table held in London with architects from insurers. The main topics were the relevance of microservices style architectures to insurance, the role of the architects in AI and InsurTech and the future role of architects at insurers. Another event that offered an interesting contrast was the inaugural London Software Architecture Conference which I'll call SACon below (Twitter feed).

Microservices

I won't fully define microservices here but briefly it’s an approach to delivering software where each service is built as it’s own application which can be scaled independently from other services.

Microservices as a way of delivering software was the default approach at the SACon. There were sessions where architects sharing stories about why sometimes you had to work with a monolith or even making the case for not having the services in discrete applications. Meanwhile at the round table the monolith was the default still with the case being made for microservices in some parts of the architecture.

There are use cases where microservices make a great deal of sense, particularly in already distributed systems where a great deal of data is being streamed between applications. Here the infrastructure of microservices and the libraries supporting the reactive manifesto such as Hysterix and Rx* (e.g. RxJava) and indeed one insurer related their use of microservices to support IoT. Others discussed using this style of approach and the tooling surrounding these architectures to launch new products and increase change throughput but in all cases these were far from replacing the core architecture.

For now microservices is not the default for insurer software but is certainly a tool in the box. An observation or two from SACon from those looking to adopt: First it doesn’t solve the question of how big a service or a component is, something architects need to discuss and refine and; Second, microservices needs a great deal of automation to make work, a topic covered in our DevOps report to be published shortly.

Architects and AI

I have a background with training and experience both in computer science, AI and machine learning. One thing that I noticed going to the analytics conferences where AI is discussed is the absence of IT representation – plenty of actuaries, MI/BI folks, marketing folks – was this a place for architects?

Most insurers present at the round table had activity within the organisation for AI. For the most part only data architects are involved in this discussion – AI being distinct from business and applications architecture for now. It’s my opinion that AI components will form part of the wider applications architecture in the future, with AI components being as common place as programmed ones.

Architects and InsurTech

Here is an area where architects can more immediately contribute in a meaningful way both in reviewing opportunities and unique capabilities from InsurTech firms and in discussing integration where acquisition rather than investment is the goal.

The challenge here of course is the age old challenge for architects – to have a seat in the discussion the architect function needs to demonstrate the value it can bring and it’s internal expertise.

Finally, one amusing discussion I had was with a few architects from startups. As I discussed legacy systems they also related seeing legacy systems in their organisations – albeit the legacy systems were 2 or 4 years old rather than 20 or 40 years old. The intriguing thing here was the reasons for them becoming legacy were the same as insurers – availability of skills, supportability and responsiveness to changing demands. It may hearten architects at insurers that start ups aren’t immune to legacy issues!

 

 

The Rise and Rise of Analytics in Insurance

The Rise and Rise of Analytics in Insurance

As noted in our prior research insurance has always been an industry that relies on advanced analytics and has always sought to predict the future (as it pertains to risk) based on the past. (For research on advanced analytics in insurers see here, here and here).

As observed in the last post here analytics, AI and automation has been a key focus of InsurTech firms but do not assume that the investment is limited to newbies and start-ups. I have for a few years now been attending and following the Strata+Hadoop conferences and others focused on advanced analytics and the broad range of tools and opportunities coming out of the big data organisations. This last week I attended a conference focused on the insurance industry and was surprised to see the two worlds have finally, genuinely overlapped – just take a look at the sponsors.

As Nicolas Michellod and I have noted in the past, insurers have already been investing in these technologies but only those that have made the effort to speak “insurance”. What the conversations at Insurance Analytics Europe (twitter feed) demonstrated was a new focus on core data science tools and capabilities. This continued the theme from DIA Barcelona (twitter) earlier in the year.

The event followed InsTech London’s meeting (Twitter) looking at data innovation and it’s opportunities for Lloyd’s, the London market and the TOM initiative. Here the focus was on InsurTech firms that would partner on analytics, would sell data or would enable non-data scientists to benefit from advances in machine learning, predictive analytics and other advanced analytics disciplines.

While this trend of democratising advanced analytics was discussed by analytics heads and CDO’s at the analytics conference the focus was much more on communicating value, surfacing existing capability and tools within the organisation and to put it bluntly, getting better at managing data.

In short – AI, Analytics, Machine Learning, Automation – these were all hot topics at InsurTech Connect and similar events but for the insurers out there – don’t assume these are purely the domain of InsurTech. Insurers are increasingly investing in these capabilities which in turn is attracting firms with a great deal to offer our industry. For those big data firms that ruled out insurance as a target market a couple of years ago – look again, the appetite is here.

As a techy and AI guy of old I am deeply enthused by this focus and excited to see what new offerings come out of the incumbent insurers and not just InsurTech.

Do have a look at the aware machine report and the blog too. We’re increasing our coverage in this area so if you have a solution focused on this space please reach out to Nicolas, Mike or myself so we can include you and for the insurers look out for a report shortly.

 

Announcing the winners of the 5th Asia Insurance Technology Awards

Announcing the winners of the 5th Asia Insurance Technology Awards
Celent and Asia Insurance Review hosted the 5th Asia Insurance Technology Awards (AITAs) at AIR’s CIO Technology Summit at Le Meridien Hotel Jakarta on 1 September 2015. The AITAs recognize excellence and innovation in the use of technology within the insurance industry. This year we received over 30 nominations from Australia, Hong Kong, Taiwan, India, Sri Lanka, Indonesia, and Pakistan; as well as the Asia Pacific divisions of global insurers. There were many impressive submissions, from which our international panel of Celent insurance analysts selected the very best to receive the six awards. The Innovation Award recognizes innovation in business models or in the use of technology. The winner was MetLife Asia. MetLife Asia implemented Advanced Data Analytics to transform big data into customer insights and to deliver a more personalized customer experience – delivering the right products and services, for the right people, at the right time. They are using these insights to inform product and services development, and to deliver sales leads to agents. The company won the award because of the innovative usage of data analytics. The IT Leadership Award honors an individual who has displayed clear vision and leadership in the delivery of technology to the business. The recipient will have been responsible for deriving genuine value from technology and has demonstrated this trait with a specific project or through ongoing leadership. The winner was Girish Nayak, Chief – Customer Service, Operations and Technology at ICICI Lombard General Insurance. ICICI Lombard implemented a business assurance project to address the ever present gap between real business uptime on the ground vs technology uptime. The firm implemented an in-house customer experience center; and deployed an infrastructure as a service model in Microsoft Azure Cloud. These initiatives generate genuine value for the business. The Digital Transformation Award honors an insurer who has made the most progress in implementing digitization initiatives. BOCG Life was the winner. BOCG Life implemented the Electronic Commerce System to provide online needs analysis and policy services. Through a transparent, direct and needs-oriented process, it facilitates prospective customers applying for multiple products they need in one go, and allows customer to adjust the offer according to their budget. The company won the award because of the way it is building trust and developing long-term relationships with customers through digital transformation. The Best Newcomer Award recognizes the best new player in the insurance technology field. The winner was CAMS Insurance Repository Services. CAMS Insurance Repository Services launched the Insurance Repository to provide e- Insurance Accounts to maintain policies as e-policies. This brings new efficiencies and benefits across the stakeholders, including Policy Holders, Insurers, Agents and the Regulator. The company won the award because they demonstrated real, unique value to the ecosystem. The award for Best Insurer: Technology honors the insurer who has made the most progress in embracing technology across the organization. The winner was RAC Insurance. RAC Insurance implemented a series of projects to digitize the business between suppliers, members and RAC Insurance. These projects include Claims Allocation, Motor Repairer Integration, and a B2C platform. The company won the award because of the way technology transformed the organization’s capability by offering an exceptional, one-touch experience for their members through online channels. Finally, the New Business Model Leveraging Mobile Applications Award recognizes the insurer who has developed a new, perhaps disruptive business model involving the innovative use of mobile technology. Max Life Insurance won the award. Max Life Insurance launched mServicing and mApp which enable digital servicing of customers, sales force and operations. The company won the award because of the use of mobile technologies to increase agent activity and engagement, enable speedy issuance of policies, and enhance business quality and operational efficiency. Be on the lookout for the 6th Asia Insurance Technology Awards in 2016. We’ll post another blog when the nomination period opens, sometime around June 2016. You can also find information on Celent’s website: http://www.celent.com/aita.

An invite to London and nothing to wear

An invite to London and nothing to wear
There are lots of cues and clues to differing cultures across the insurance industry and it’s IT neighbour – one of the most obvious is dress code or at least communal agreement on how one should dress. For a chap in London it should be relatively easy, as the character Harry Hart put it in the film Kingsman, “The suit is the modern gentleman’s armour.” However, recent changes and external influences in London have left me in something of a wardrobe quandary. For example – the data scientist community and the digital community. I went to the first Strata event in London in my usual suit and tie and swiftly realised that I looked like I a fish very much out of water. Here jeans, t-shirts and the odd tattoo were the order of the day. My most recent visit to the conference I managed to correct my attire although didn’t acquire new tattoos just for the conference (perhaps next year). Oliver Werneyer’s observation at our event in February this year that one needs a good beard to fit in with the start up crowd is also well founded. Also in London we have Lloyd’s of London with a strict dress code and a requirement for a tie to be worn at all times. More Kingsman territory, clearly one can’t dress for both communities on the same day. In between we have an increasingly relaxed view of the suit attire or even simply trousers and shirt. Despite having a pretty good collection of ties these are now largely optional (although I still generally carry one around as wearing them varies by client and frankly I quite like wearing a tie to a meeting). What I don’t have of course is a pocket square – something I rarely have seen adopted before this year (perhaps I wasn’t paying attention) but I’m increasingly seeing a square used to add a splash of colour in the absence of a tie. Thus, we have the title of this post – I have nothing to wear! Fortunately, London is unlikely to see the weather required for hawaiian shirts and shorts to become the order of the day (albeit I may have something that might fit that bill should it come to pass). Circling back to culture though, the need to blend these clearly different and shifting cultures together in one organisation is crucial in a modern insurer. Aviva has gone to the extent of creating a digital garage in Shoreditch – the heart of the jeans wearing community, if I may use such a broad brush – to draw in talent to the organisation. Hiscox too has been going to great pains to attract the right talent, along with many other insurers in London seeking to bridge these cultures. Are you allowing for a varied culture in your organisation? How flexible are you in dress code and working practices across different communities? Have you ever set to preparing for a meeting and realised you simply have nothing to wear? Would love to hear your stories on changing insurance, if only so I know it’s not just me.  

Ace buys Chubb: what it means for insurance technology

Ace buys Chubb: what it means for insurance technology
Today’s blockbuster announcement of Ace buying Chubb will have a lot of industry ramifications—some of which will play out in the IT sphere. No doubt there has already been an IT assessment element in each insurer’s due diligence efforts. Between now and the effective date of the merger, there will be a lot of planning focused on:
  • Efficiencies and platform rationalization–aka “let’s figure out what is the right number of core systems, which core systems will be the survivors, and how data conversion and integration will work”
  • Cloud, SaaS, data management/stores, and analytics
  • Professional service and SI support capabilities that can scale to the new Chubb
  • Which systems will best support a digital roadmap
Some seemingly redundant systems may survive—at least over a 1 to 3 year period. For that to happen, the business (and/or various geographies’ compliance) requirements of the operating units using these system will be too divergent or too difficult to quickly build into a single surviving system. All this reinforces the reigning market message to insurance technology firms. If you want to be around in 10 years:
  • Design highly configurable and agile systems that feature ease of integration
  • Have enough scale to meet the needs of bigger and bigger insurer customers—grow, merge, or wither
 

Open source, analytics and the pace of change

Open source, analytics and the pace of change
I love spotting ironies such as how this years Strata | Hadoop World conference (the UK one) spent more time discussing Apache Spark and whether it was a successor to Hadoop or another tool in the box than it did discussing Hadoop and it’s applications. It was great to see members of the insurance industry there amongst the retailers and banks as well. “But wait?!?!!” I hear you say, “Hadoop isn’t all that old is it?” Herein lies the great challenge for the CIO faced with requests for open source tools. These are dynamic, social projects without the same stickiness as those legacy systems insurers spend time worrying about. Not only do users / consumers / fans of open source software shift between projects but the contributers / developers do too. With the rising use of tools like R, Python, Linux, GIT, Hadoop, Spark, Docker, Capistrano and all manner of wacky projects on the go and being adopted by insurers how should a CIO respond? Prohibition tends to lead to shadow IT and surprises down the line far more unwelcome than managing some new software. The key advice is to understand these types of projects can be more transient than other enterprise software. Experiment with them but be careful of expensive, enterprise installations that are hard to extract later down the line. In truth insurer adoption of some of these technologies will outlive the fashion for them but it still requires planning for their removal or worst case, their ongoing support. I promised analytics in the title too didn’t I? Well Spark is all about real time analytics and is having an interesting impact in the machine learning and predictive modelling space. It gets around some of the issues with interacting with Hadoop while still delivering performance. With open source projects survival of the fittest is the order of the day, far more so than in classic insurance software markets. Hadoop has it’s place, with many insurers globally investing in it.We will see new fashions in analytics approaches and more opensource tools I’m sure. Some will follow the Dodo. For those interested in Hadoop have a look at my report from 2011, when Hadoop was new and cutting edge. It seems it requires an update.

The rise and fall (and rise) of Artificial Intelligence

The rise and fall (and rise) of Artificial Intelligence
Artificial intelligence has been around nearly as long as humans have been able to think about themselves, about thought and what they do. Empathy is wired into us – some more than others but we are all capable of thinking from another’s point of view. This capacity leads us to anthropomorphize things that aren’t human, to imbue things in our daily lives with human qualities like moods, characteristics and personality. When we build puppets, robots, models that look sort of human it is easy to for us to assign it with greater power, ability and promise than is really there. For marketers in other fields, to have consumers attribute their products with ‘magical’ properties would be a dream come true but for artificial intelligence it is a nightmare – one the industry has expended funds marketing against. Artificial intelligence has delivered many great tools which today we take for granted. Our phones listen to us and understand our requests in the context of our calendar, our camera’s recognise faces and social networks tell us who those faces belong to, machines translate words from one language to another (although don’t get the translations tattooed just yet) and the list goes on. We chuckle at these mistakes these learning and adaptive systems make, we see the huge strides and investment and we expect a new human like intelligence to emerge in the short term. Around the middle of every decade since the 60’s there has been a peak in excitement for AI, a frustration with it’s lack of progress, and a reduction of funding or AI winters as they are called. In the eighties it was LISP machines, in the nineties it was expert systems. Now in the twenty-tens (I thought it was teenies but that’s a kids show apparently) we are seeing a resurgence of AI, a blending of machine learning, predictive modelling and cognitive computing along with self driving cars. This raises some rare and interesting questions:
  • Are we headed for a new AI winter?
  • Or an AI apocalypse?
  • Also, will I still be cleaning my home in 2020?
It is certainly true to say the set of tasks we can expect software and physical computing systems to do is vastly increased compared to just a decade ago, and massively so since the 60s. Doing all the things humans can do and living in our society, empathising and understanding us in that broad context is still well beyond computers – but engaging with us in specific, well-defined domains such as about our calendar or what we would like to buy from the shop is well within their grasp today. Previously difficult tasks such as searching a database for information, reading that from a screen and keying it into another screen is now entirely possible – see the earlier blog post on bots. Having a drone fly itself around an obstacle to reach an objective is still very hard. Having a vehicle drive itself on the road is in fact easier, albeit most humans don’t benefit from lidar sensors, ultrasonics and eyes in the back of their head (alright, bumper). It is good to see AI on the rise again – I loved the topic ever since getting into programming and getting involved in a cognitive psychology course some years ago. I recall writing an expert system in Pascal back in the 90s. I am concerned, as the insurance industry should be, by a new AI winter. Self driving cars and vehicles have the potential to make the roads safer for all. We will when we see them, imbue them with more power than they have – this is human nature. We will, in the not too distant future, hear people say things like, “the car likes to give cyclists a lot of room on the road” or “the car prefers to take this corner at a fair speed” – imbuing a complex machine with sensors and programming with preferences, desires and likes – human qualities. When the first death comes we will ask how could it do such a thing. When an automated car is put in a position where it must decide between a set of actions – each leading to injury, we will hear people discuss why it chose to do what it did, people may say, “it did the best it could” or worse, “no person would ever have done that, this is why machines shouldn’t be able to choose.” The latter of course revealing the human construct, an unspoken contract – our expectation that smart or intelligent systems will operate like us, share our values, our culture, that we can predict their actions in our context. This is the greatest threat to AI and always has been – the expectation, the contract that the new intelligence will be like human intelligence. Some winters are due in part to that contract being broken, to these systems not living up to the expectation and making inhuman mistakes. There are a set of tools available now that are not intelligent but they are smart and they are powerful. We would be remiss in our duty to our customers and shareholders if we do not leverage them. Manage expectations about these powerful tools and understand the very real limits that exist on them. If we can do this we may benefit from the AI boom and avoid another AI Winter. Will we see an AI apocalypse? Ironically it’s not the human like intelligence that may be our greatest threat but simpler intelligences. A human like intelligence could empathise, could act in acordance with values and could be relatively predictable (in a human way). There are many stories across science fiction of smart robots that act like insects and replicate, in fact that only make copies of themselves, that pose a great threat to any civilisation. They are not intelligent, they don’t want to kill off all life in the galaxy – they just turn all the available resources into copies of themselves which would have that effect. We are much closer to building that threat frankly (with drones, 3d printers, etc.), than a super intelligence that decides all human life is worthless. For now though – I expect these things to stay firmly in the space of science fiction. I include this discussion here because it does demonstrate a key difference between smart with unintended consequences and ‘intelligent’ – a lesson worth bearing in mind for those adopting AI. Finally – will we see robots cleaning our homes by 2020? Well roomba is out there and sort of does that. Stairs and steps are still a huge challenge to robots. Frankly differentiating furniture, pets, clutter, magazines, rubbish, dust and recycling in a moving environment is still a very complex issue. As in insurance, I think smart things will make cleaning easier and assist those who invest but there’ll be a role for human intelligence in ensuring the pets aren’t recycled and the customer ultimately gets the service they expect.

A Recipe for Digital Innovation

A Recipe for Digital Innovation
At each of the five Celent Innovation Roundtables held in the last several months, innovation practitioners consistently identify culture change as a significant success factor. A particular challenge, poor communication between technologists and their business partners, is often cited as a barrier. The Second Machine Age by MIT professors Erik Brynjolfsson @erikbryn and Andrew McAfee @amcafee offers some help. Their explanation of digital innovation made a big impression on me as the clearest description that I have found so far.  The approach is simple: “digital information….is built on multiple layers”. It is a “recipe” of different automation solutions mixed together. That is, look at a list of digital technologies, pick a few and combine them in unique ways so that they work together, and deliver new value. This description led me to revisit some Celent insurance innovation case studies and rethink how to best explain them.  The first, the AXA claims example (Visualizing the London Riots at AXA UK, http://www.celent.com/reports/visualising-london-riots-axa-uk), outlined how the insurer combined data from public police records, media reports, and their internal systems to predict which of their insureds might suffer a loss during the multi-day rioting in the U.K. in 2011. AXA “layered” successive sources of digital data, then added some analytic algorithms to produce a new and valuable tool designed to proactively identify at-risk insureds (mainly small businesses that were exposed to looting). All of these technologies existed on their own, in isolation, until they were combined to yield new insights which helped avoid losses. The second study is from Tokio Marine & Nichido Fire Insurance Co., Ltd. They were recognized as a Celent Model Insurer for their One Time Insurance product (Model Insurer 2012: Case Studies of Effective Technology Use in Insurance http://www.celent.com/reports/model-insurer-2012-case-studies-effective-technology-use-insurance). They combined geo-location, text messaging, and data prefill services to deliver real-time insurance offers to subscribers. As a prospect drives to the airport, their mobile phone receives a text from the insurer with an offer for travel insurance. Similarly, texts are sent as golfers arrive for their tee times, skiers approach the lifts, etc. It is the combination, or layering, of these technologies in a unique manner that creates the innovative service. The value of this explanation is not only academic. Layering strikes me as a useful tool to explain how all of this “digital stuff” can fit together. The recipe and layering metaphors succinctly describe digital in non-technical, accessible terms. It can be used with any audience to illustrate how the sum of the parts can be greater than the whole. I also see value in using layering to generate new ideas. My thought is that, in an interactive session, a group of participants can create a list of technologies, data sources, etc. and then brainstorm different combinations from them. Our continuing research illustrates that there is no one prescription for innovation, but there are guideposts to follow.  The use of the layering metaphor to improve communication and as a technique for brainstorming is one such guide.