Dawn of the machine-to-machine age and its implications for insurance | Quantumrun

Dawn of the machine-to-machine age and its implications for insurance

Home / Business forecasts
By Syed Danish Ali
Sep 03, 2018,  7:10 PM

Machine-to-machine technology (M2M) essentially involves sensors in an Internet of Things (IoT) environment where they send data wirelessly to a server or another sensor. Another sensor or server uses Artificial Intelligence (AI) to analyze the data and act upon the data automatically in real time. The actions can be anything like alerts, warning, and change in direction, brake, speeding, turning, and even transactions. As M2M are exponentially increasing, we will soon see reinventing of whole business models and customer relationships. Indeed, the applications will only be limited by the imagination of businesses.

This post will explore the following:

  1. Overview of key M2M technologies and their disruptive potential.
  2. M2M transactions; a whole new revolution where machines can directly transact with other machines leading to the machine economy.
  3. AI’s impact is what is leading us to M2M though; big data, deep learning, streaming algorithms. Automated machine intelligence and Machine teaching. Machine teaching is perhaps the most exponential trend of the machine economy.
  4. Insurance business model of the future: Insuretech startups based upon blockchain.
  5. Concluding remarks

Overview of key M2M technologies

Imagine some real-life scenarios:

  1. Your car senses your travel journey and buys insurance on an on-demand basis by the mile automatically. A machine buys its own liability insurance automatically.
  2. Wearable exoskeletons giving law enforcement and factory works superhuman strength and agility
  3. Brain-Computer interfaces merging with our brains to create super-human intelligence (for example, Neural Lace of Elon Musk)
  4. Smart pills digested by us and health wearables directly assessing our mortality and morbidity risks.
  5. You can get life insurance from taking a selfie. The selfies are analyzed by an algorithm that medically determines your biological age through these images (already being done by Chronos software of startup Lapetus).
  6. Your fridges understand your regular shopping and stocking habits and find that some item like milk is getting over; so, it buys milk through online shopping directly. Your fridge will be continuously re-stocked based on your most common habits. For new habits and non-usual, you can continue to independently buy your items and stock it in the fridge as usual.
  7. Self-driving cars interact with each other on the smart grid to avoid accidents and collisions.
  8. Your robot senses that you are getting more upset and depressed lately and so it tries to cheer you up. It tells your health coach bot to increase content for emotional resilience.
  9. Sensors sense an upcoming burst in the pipe and before the pipe bursts, sends a repairman to your home
  10. Your chatbot is your personal assistant. It does shopping for you, senses when you need to buy insurance for let’s say when you are travelling, handles your daily chores and keeps you updated on your daily schedule that you have made in collaboration with the bot.
  11. You have a 3D printer for making new toothbrushes. The current smart toothbrush senses that its filaments are about to get worn out so it sends a signal to the 3D printer to make new filaments.
  12. Instead of bird swarms, we now see drone swarms flying off carrying out their tasks in collective swarm intelligence
  13. A machine plays chess against itself without any training data and beats just about everyone and everything (AlphaGoZero already does this).
  14. There are countless real-life scenarios like these, limited only by our imagination.

There are two meta-themes arising out of M2M technologies: prevention and convenience. Self-driving cars can eliminate or radically reduce accidents as the majority of car accidents are caused by human errors. Wearables can lead to a healthier lifestyle, smart home sensors pipe bursts and other issues before they occur and rectify them. This prevention decreases morbidity, accidents and other bad events. Convenience is an over-arching aspect in that most everything happens automatically from one machine to another and in few remaining cases, it is augmented with human expertise and attention. The machine learns what it is programmed to learn on its own using data from its sensors about our behaviors over time. It happens in the background and automatically to free up our time and efforts onto other more human things like being creative.

These emerging technologies are leading to changes in exposures and have huge impact on insurance. A large number of touch points are made where the insurer can engage with the customer, there is less focus on personal coverage and more on commercial aspect (like if self-driving car malfunctions or gets hacked, home assistant gets hacked, smart pill poisons instead of providing real-time data to dynamically assess mortality and morbidity risks) and so on.  The frequency of claims is set to radically decrease, but the severity of claims can be more complex and difficult to assess as various stakeholders will have to be taken on board to assess the damages and to see how the share of loss coverage varies in proportion to faults of different stakeholders. Cyber hacking will multiply leading to new opportunities for insurers in the machine economy.  

These technologies are not alone; capitalism cannot exist without constantly revolutionizing technology and thereby our human relations with it. If you need more awareness of this, see how algorithms and technology is moulding our mentalities, thinking attitudes our behavior and actions and see how rapidly evolving all technology is.  What’s surprising is this observation was made by Karl Marx, someone who lived in 1818- 1883 and this shows that all the tech in the world is no substitute for deep thinking and erudite wisdom.

Social changes go hand in hand with technological changes. Now we are seeing peer to peer business models with a focus on social impact (Lemonade for example) instead of only making the rich richer. The sharing economy is boosting the use of technology as it provides access (but not ownership) to us on an on-demand basis. The millennial generation is also very different from previous generations and we have only started waking up to what they demand and how they want to shape the world around us. The sharing economy can mean that machines with their own wallets can perform services on an on-demand basis for humans and transact independently.

M2M financial transactions

Our future customers will be machines with wallets. A cryptocurrency called “IOTA (Internet of Things Application)” aims to propel the machine economy into our everyday reality by allowing IoT machines to transact to other machines directly and automatically and this will lead to rapid emergence of machine-centered business models.

IOTA does this by removing blockchain and instead adopting ‘tangle’ distributed ledger which is scalable, lightweight and has zero transaction fees which mean that micro-transactions are viable for the first time. The key advantages of IOTA over current blockchain systems are:

  1. To allow a clear idea, blockchain is like a restaurant with dedicated waiters (miners) that bring you your food. In Tangle, it’s self-service restaurant where everyone serves themselves. Tangle does this through the protocol that person has to verify his/her previous two transactions when doing a new transaction. Thus miners, the new middleman building up immense power in blockchain networks, are altogether made useless through Tangle. The promise of blockchain is that middlemen exploit us whether they be the government, the money-printing banks, the various institutions but another class of middlemen ‘miners’ are becoming quite powerful, especially Chinese miners leading to a concentration of huge power in a small number of hands. Bitcoin mining takes as much energy as electricity produced by more than 159 countries so it is a huge waste of electricity resources as well because huge computing hardware are required to crack complex crypto mathematical codes to validate a transaction.
  2. As mining is time-consuming and expensive, it doesn’t make sense to perform micro or nano transactions. Tangle ledger allows transactions to be validated in parallel and require no mining fees to vitally allow the IoT world to conduct nano and microtransactions.
  3. Machines are ‘unbanked’ sources in today’s time but with IOTA, machines can generate income and become an economically viable independent unit that can purchase insurance, energy, maintenance etc. on its own. IOTA provides “Know Your Machine (KYM)” through secure identities like the banks have currently Know Your Customer (KYC).

IOTA is a new breed of cryptocurrencies that aim solve problems that previous cryptos were not able to solve. The “Tangle” distributed ledger is a nick-name for a Directed Acyclic Graph as shown below: 

Directed Acyclic Graph is a cryptographic decentralized network that is supposedly scalable till infinity and resists attacks from quantum computers (which are yet to be commercially fully developed and used in mainstream life) through using a different form of encryption of hash-based signatures.  

Instead of becoming cumbersome to scale, the Tangle actually speeds up with more transactions and gets betters as it scales up instead of deteriorating. All devices using IOTA are made part of Node of the Tangle. For every transaction done by the node, node 2 must confirm other transactions. This way there is twice as much capacity available as the need to confirm the transactions. This anti-fragile property in which tangle improves by chaos instead of worsening due to chaos is a key advantage of the Tangle.

Historically and even presently, we induce trust upon transactions by recording their trail to prove the transactions’ origin, destination, quantity and history. This requires huge time and efforts on part of many professions like lawyers, auditors, quality inspectors and many support functions. This, in turn, causes humans to kill their creativity by becoming number-crunchers doing manual verifications to and fro, causes transactions to be expensive, inaccurate and expensive. Too much human suffering and Dukkha has been faced by many humans doing monotonous repetitive jobs just to create trust in these transactions. As knowledge is power, important information is kept hidden by those in power in determent to the masses. The blockchain is allowing us to potentially ‘cut through all of this crap’ of the middlemen and give power to the people through technology instead which is the chief goal of the fourth industrial revolution.

However, current blockchain has its own set of limitations regarding scalability, transaction fees and computing resources that are required to mine. The IOTA does away with blockchain altogether by replacing it with ‘Tangle’ distributed ledger to create and verify transactions.  The purpose of IOTA is to act as a key enabler of the Machine Economy which, so far, has been restricted due to limitations of current cryptos.

It can be reasonably forecasted that many cyber-physical systems will emerge and be based upon Artificial Intelligence and IoT such as supply chains, smart cities, smart grid, shared computing, smart governance and healthcare systems. One country with very ambitious and aggressive plans to become well known in AI beside the usual giants of USA and China is the UAE. UAE has so many AI initiatives like it has shown drone police, plans on driverless cars and hyperloops, governance based on blockchain and even has the first state minister in the world for Artificial Intelligence.

The quest for efficiency was the quest which first drove capitalism and now this very quest is now working to end capitalism. 3D printing and sharing economy are radically lowering costs and upgrading efficiency levels and the ‘Machine Economy’ with machines with digital wallets is the next logical step to greater efficiency. For the first time, a machine will an economically independent unit earning income by physical or data services and spending on energy, insurance and maintenance all on its own. On-demand economy will boom because of this distributed trust. 3D printing will radically bring down the cost of making materials and robots and economically independent robots will soon start giving services on an on-demand basis to humans.

To see the explosive impact it can have, imagine replacing centuries old Lloyd’s insurance market. A startup, TrustToken is trying to create a trust economy to carry out transactions USD 256 trillion, which is the value of all real-world assets on earth. The current transactions take place in outdated models with limited transparency, liquidity, trust and a lot of problems. Carrying out these transactions using digital ledgers like blockchain is far more lucrative through the potential of tokenization. Tokenization is the process through which real world assets are converted into digital tokens.  TrustToken is making the bridge between digital and real worlds through tokenizing real world assets in a way that is acceptable in the real world too and is ‘legally enforced, audited and insured’. This is done through the creation of ‘SmartTrust’ contract that guarantees ownership with legal authorities in the real world, and also implements any necessary action when contracts are broken, including reposing, charging criminal penalties and much more. A decentralized TrustMarket is available for all stakeholders to gather and negotiate the prices, services and TrustTokens are the signals and rewards parties receive for trustworthy behavior, to create an audit trail and to insure the assets.

Whether TrustTokens are able to carry out sound insurance is a matter up for debate but we can already see this in the centuries old Lloyd’s market. In Lloyd’s market, buyers and sellers of insurance and underwriters gather together to carry out insurance. An administration of Lloyd’s funds monitors their various syndicates and provides capital adequacy to absorb the shocks that come from insuring too. TrustMarket has the potential to become the modernized version of Lloyd’s market but it is too early to determine its precise success. TrustToken can open up the economy and create better value and lesser costs and corruption in real world assets, especially in real estate, insurance and commodities that create too much power in the hands of the very few.

The AI part of the M2M equation

Much ink has been spelt on AI and its 10,000+ machine learning models that have their own strengths and weaknesses and are allowing us to uncover insights that were hidden from us before to radically improve our lives. We will not describe these in detail but focus on two areas of Machine Teaching and Automated Machine Intelligence (AML) as these will allow IoT to transform from isolated bits of hardware to integrated carriers of data and intelligence.

Machine teaching

Machine teaching, is perhaps the most exponential trend that we are seeing which can allow M2M economy to prop up exponentially from humble beginnings to become a dominant feature of our everyday lives. Imagine! Machines not only transacting with each other and other platforms like servers and humans but also teaching each other. This has already happened with the Tesla Model S’s autopilot feature. The human driver acts as the expert teacher to the car but the cars share these data and learning between themselves radically improving their experience in extremely short time. Now one IoT device is not an isolated device that will have to learn everything from scratch on its own; it can leverage the mass learning learned by other similar IoT devices worldwide as well. This means that intelligent systems of IoT trained by machine learning are not just becoming smarter; they are getting smarter faster over time in exponential trends.

This ‘Machine Teaching’ has huge advantages in that it lowers the training time required, bypasses the need to have massive training data and allows machines to learn by themselves to improve the user experience. This Machine Teaching can be sometimes collective like self-driving cars sharing and learning together in sort of a collective hive mind, or it can be adversarial like two machines playing chess against itself, one machine acting as the fraud and the other machine as the fraud detector and so on. The machine can also teach itself by playing simulations and games against itself without the need of any other machine. AlphaGoZero has done exactly that. AlphaGoZero did not use any training data and played against itself and then defeated the AlphaGo which was the AI that had defeated the world’s best human Go players (Go is a popular version of Chinese chess). The feeling that chess grandmasters had of watching AlphaGoZero play was like an advanced alien super-intelligent race playing chess.

The applications from this are staggering; hyperloop (very fast train) based tunnel pods communicating with each other, autonomous ships, trucks, whole fleets of drones running on swarm intelligence and the living city learning from itself through smart grid interactions.  This along with other innovations occurring in the fourth industrial revolution of Artificial Intelligence can eradicate current health problems, many social problems like absolute poverty and allow us to colonize Moon and Mars.

Aside from IOTA, there is also Dagcoins and byteballs that don’t require blockchain. Both Dagcoins and byteballs are again based on DAG Directed Acrelic Graph just like ‘tangle’ of IOTA is. Similar advantages of IOTA apply roughly to Dagcoins and byteballs as these all overcome current limitations of blockhain.

Automated machine learning

There is of course a broader context to automation where almost every field is suspect to and no one is free from this fear of AI apocalypse. There’s also a brighter side of automation where it will allow humans to explore ‘play’ instead of work only. For a comprehensive coverage, see this article on futurism.com

Despite the hype and glory associated with quantitative modelers like data scientists, actuaries, quants, and many others, they face a conundrum which automated machine intelligence sets out to solve. The conundrum is the gap between their training and what they should be doing compared to what they actually do. The bleak reality is most of the time gets taken by monkey work (work that any monkey can do instead of an intellectually trained and competent human being) like repetitive tasks, number crunching, sorting out data, cleansing data, understanding it, documenting the models and applying repetitive programming (being spreadsheet mechanics too) and good memory to remain in touch with all of that mathematics. What should they be doing is being creative, producing actionable insights, talking with other stakeholders to bring about concrete data-driven results, analyzing and coming up with new ‘polymath’ solutions to existing problems.

Automated machine intelligence (AML) takes care to reduce this huge gap. Instead of hiring a team of 200 data scientists, a single or few data scientists using AML can utilize fast modeling of multiple models at the same time because most of the work of machine learning is already automated by AML like exploratory data analysis, feature transformations, algorithm selection, hyper parameter tuning and model diagnostics. There are a number of platforms available like DataRobot, Google’s AutoML, Driverless AI of H20, IBNR Robot, Nutonian, TPOT, Auto-Sklearn, Auto-Weka, Machine-JS, Big ML, Trifacta, and Pure Predictive and so on AML can compute dozens of suitable algorithms in the same time to find out optimum models according to pre-defined criteria. Whether they are deep learning algorithms or streaming algorithms, all are automated neatly to find the optimum solution which is what we are actually interested in.

Through this way, AML frees up data scientists to be more human and less cyborg-Vulcan-human calculators.  Machines are delegated to what they do best (repetitive tasks, modeling) and humans are delegated to what they do best (being creative, producing actionable insights to drive business objectives, creating new solutions and communicating them). I cannot say now that ‘wait first let me become a phD or expert in Machine Learning in 10 years and then I will apply these models; the world moves too fast now and what’s now relevant becomes outdated very quickly. A fast paced MOOC based course and online learning makes far more sense now in today’s exponential society instead of the fixed-one-career-in-life that previous generations are used to.

AML is necessary in the M2M economy because algorithms need to be developed and deployed in the ease with little time. Instead of algorithms requiring too many experts and they taking months to develop their models, AML bridges the time gap and allows for enhanced productivity in applying AI to situations which was unthinkable before.

Insuretechs of the future

To make the process further seamless, agile, robust, invisible and as easy as a child playing, blockchain technology is used with smart contracts that execute itself when the conditions meet. This new P2P insurance model is doing away with traditional premium payment using instead a digital wallet where every member puts in their premium in an escrow-type account only to be used if a claim is made. In this model, none of the members carry an exposure greater than the amount they put into their digital wallets. If no claims are made all digital wallets keep their money. All payments in this model are done using bitcoin further reducing transaction costs. Teambrella claims to be the first insurer using this model based on bitcoin. Indeed, Teambrella is not alone. There are many blockchains based startups targeting peer to peer insurance and other areas of human activity. Some of them are:

  1. Etherisc
  2. Insurepal
  3. AIgang
  4. Rega Life
  5. Bit Life and Trust
  6. Unity Matrix Commons

Thus, a lot of crowd wisdom is utilized in this as the insurer ‘Learns from the people, plans with the people, Begins with what they have And Builds on what they know’(Lao Tze).

Instead of an actuary maximizing profit for the shareholders, sitting isolated from ground realities, lacking skin in the game, and have far less access to awareness (i.e., data) of people relative to their peers, this peer to peer empowers the crowd and taps in into their wisdom (instead of wisdom from books) which is far better. There are also no unfair pricing practices here like rating based on gender, pricing optimization which charges you higher if you are less likely to shift to another insurer and vice versa. The giant insurer cannot know you more than your peers, it’s as simple as that.

These same peer-to-peer insurance can be carried out on non-blockchain based distributed ledgers too like IOTA, Dagcoins and Byteballs with additional technological benefits of these new ledgers over current blockchain. These digital tokenization startups have the promise to radically reinvent business models where transactions, pooling and just about anything gets done for the community and by the community in an automated fullytrustworthyy manner with no oppressive middlemen like governments, capitalist businesses, social institutions and so on. Peer to Peer Insurance is just one part of the whole program.

Smart contracts have built-in conditions with them which are automatically triggered when the contingency happens and claims get paid instantly. The huge need for labor force with high qualifications but essentially doing clerical work is removed altogether to build a sleek autonomous organization of the future. The oppressing middleman of ‘shareholders’ are avoided which means that consumer interests are acted upon by providing convenience, low prices and good customer support. In this peer to peer setting, the benefits goes to the community instead of the shareholder. IoT provides the main source of data to these pools to develop protocols when to release claim payment and when not to.  The same tokenization means that anyone anywhere can have access to the insurance pool instead of being limited by geography and regulations.

Impact (ONLY use the 'Paste From Word' button to safely copy and paste text from a Word doc) 

The scenario pictured here of M2M transactions and the machine economy will look far off and distant to many, especially as current realities of insurers is still the same as it was 200 years ago. The machine economy is still nascent and emerging, but exponential results can mean that it becomes a large part of our lives very soon. The current insurers are yet to embrace emerging trends with a more immediate impact like insuretech that don’t use blockchain like Trov, Lemonade, Verifly and many others. We have yet to see the full-blown utilization of IoT and M2M but we are certainly in the stages of seeing its dawn and emergence. 

Forecasted start year: 
2026 to 2030


Load comments