In my last blog about AoE we introduced several topics to demonstrate AI’s impact including the size of the financial services marketplace, the meaning of machine learning as the basis of AI, the proliferation of Chatbots and how AI is the inflection point for financial services and AI’s wide coverage, usage, and proliferation of tools. In part two, we continue this discussion.
AI is driving the enterprise
AI is driving new innovation business strategy due to machine learning integration into all nine areas of Alex Osterwalder’s Business Model Canvas that is the foundation of agile enterprises:
- Key areas (partners, activities, resources)
- Value proposition(s)
- Customer (relations, segments)
- Outreach channels
- Financial (costs, revenue)
In particular, machine learning can significantly increase delivery on the enterprise’s unique or differentiating value propositions of:
- First mover on something new
- Outstanding capabilities
- Ease in tailoring to individual needs
- Speed in initial usage
- Desired brand and associated high status by association
- Best price
- Lower operational costs
- Risk lowering and mitigation
- Ease of access and continuing use
Agility and innovation is creating complete restructuring of the biggest enterprises, GE being a prime example with efficiency improvements at a fraction of the cost.
Types of innovation are extended to include: products, services, process, organizational model, business model and the 9 areas of the business model canvas, social-mediated and with machine learning integration in all innovation types.
There are daily announcements in AI centered on an AI-first strategy building on previous generations of cloud first and mobile first. This is another switch in business focus and one financial services needs to examine.
The growing signs of AI planning and implementation
- A Hong Kong VC fund, Deep Knowledge Ventures, assigning an AI to its board for decision making.
- Baidu, China’s dominant search company, announcing an AI-based StockMaster App that analyses news, markets predicting sectors, stocks or markets changes.
- Telefonica and BigML using AI to select start-ups for funding.
- Accelerating this year, we see growing use of AI-based robo advisors in wealth management and the proliferation of intelligence agents and chat bots. Robo-advisors removing the need for brokers, generating higher returns with little cost/lower fees, reducing minimums, and growing alignment/anticipation with consumer needs/wants/behaviors.
- There is AI’s increasing implementation in consumer financial and healthcare apps pre-emptively guiding our daily financial and health lifestyle choices.
- McKinsey indicating that 58 per cent of US jobs can be automated with AI-based natural language processing working at average human levels.
- In insurance eliminating the agent and with better service and at reduced costs.
How about predicting major global changes which impact financial services?
There is NELL, the Never-Ending Language Learner which is consuming the web with more than 50 million items learned.
Ultimately, this becomes an uber prediction tool much like Bing Predicts from Microsoft and Google prediction technology and with machine learning tools easily usable for Fintech and financial services.
At a recent board meeting, a scientist commented that Google had long predicted Trumps success in the US political race.
When I keynoted on Megatrends in Korea, another speaker hosted by the Ministry of Science, ICT and Future Planning talked about a hybrid delphi system Korea is using to predict and then act upon future trends with over 80 per cent accuracy. It’s a combination of human/machine collaboration and a catalyst where they are trying to change to an algorithm based economy from component manufacturing and with over 4.5 per cent of their GDP going to R&D, the highest in the world.
What of the future?
The impact of AI is so profound and so widespread that Bill Gates in June provided a recommended must read of, The Master Algorithm by Pedro Domingos. Pedro sits on the FSR Technology Advisory Council and his book describes the five tribes of machine learning:
- Symbolists–Fill in gaps in existing knowledge
- Connectionists–Emulate the brain; this is Deep Learning
- Evolutionists–Simulate evolution
- Bayesians–Systematically reduce uncertainty
- Analogizers–Notice similarities between old and new
This year, I keynoted at ICSE, the world’s largest software engineering conference funded by the National Science Foundation, the research arms of the major technology companies and the four top science organizations in software engineering. Earlier from Pedro, I received his five top AI megatrends which I outlined at ICSE. I am reproducing them here due to their impact on financial services and due to the pronounced impact on the international audience of top researchers:
1.The transition from computers that are programmed by us to computers that learn on their own. This is enabled by big data, and in turn enables the personalization of everything, from medicine to shopping, and the increasing automation of every function in an organization.
2.The automation of scientific discovery. Increasingly, each step of the scientific method, from gathering data to formulating hypotheses, is carried out by computers. This enables, for example, new drugs to be discovered at a much faster rate than before.
3.The replacement of white-collar workers by machines, not just blue-collar ones. Routine intellectual work can increasingly be done by AI; what’s hard to replace is physical dexterity, common sense, and integrative intelligence.
4.The transition from deterministic to probabilistic computing. From hardware to software, rigidly deterministic computations are giving way to probabilistic ones, enabling faster, cheaper, lower-power, larger-scale, more ubiquitous, more flexible, data-driven information systems.
5.The rise of evidence-based X, where X includes medicine, policy-making, development aid, and ultimately all important societal decisions. Instead of guesswork and mixed results, we have controlled trials that quickly weed out what doesn’t work from what does.
All of this creates the “AI of Everything (AOE) planetary machine learning mesh driving a financial services digital quake and inflection point.” Moreover, the discussion on AI at the January 2017 FinTech Ideas Festival is one to watch.