The Future of Applications
There are a number of reasons “why now” is the right time for companies and entrepreneurs to build intelligent applications, but the key reasons that we see are:
• The world’s data doubling every two years
• The availability of massive computational power and low-cost storage to feed machine learning models
• The ease of use with which developers can take advantage of machine learning techniques
• The adoption of microservices as a development paradigm for applications
• The proliferation of platforms on which to develop applications, and in particular platforms based on "natural user interfaces" like messaging and voice
• And, most importantly, the widespread consumer expectation for all applications to serve more personalized and relevant content
Whether your company is building a consumer-facing application, an enterprise SaaS app, or the building blocks for other machine learning services, here are key things that we have heard from machine learning experts on how to build successful intelligent applications:
1. Build for an intelligent future by creating a data moat
The famous computer science and machine learning expert Andrew Ng describes machine learning and AI as the “new electricity.” By replacing steam-powered machines with electric machines, the world was able to make massive progress in the fields of transportation, manufacturing, agriculture, healthcare, and so on. AI will be the next force that drives a similar level of productivity gain, and to power all of these AI-driven systems, the core “fuel” is no longer coal or natural gas. It is data. In order for companies to remain competitive in the future, they will need to either gain access to or create their own proprietary data sets.
2. Understand build vs. buy decisions in the context of your company’s competitive advantage
Companies building intelligent applications today fall into two categories broadly – ones that are building some form of ML/AI technology or ones that are using ML/AI technologies in their application.
Every successful, new application built today and in the future, will be an intelligent application
There is a tremendous amount of innovation that is currently happening in the building block services that include both data preparation services, learning services, and models-as-a-service providers. With the advent of microservices and the ability to seamlessly interface with them through REST APIs, there is an increasing trend for the learning services and ML algorithms to be used and re-used as opposed to having to be re-written from scratch over and over again. For example, Algorithmiaruns a marketplace for algorithms that any intelligent application can use as needed. Combining these algorithms and models with a specific slice of data within a particular vertical is what we call micro-intelligence that can be seamlessly incorporated into applications.
3. Ensure that trust and transparency are a central part of your product
Several high-profile experiments with ML and AI came into the spotlight in the last few years. Examples include Google DeepMind AlphaGo, Facebook M, and the introduction of a number of different autonomous vehicles. For most consumers, understanding the ‘why’ behind the ‘what’ is a critical component of working with artificial intelligence. A doctor or a patient will not be happy with a diagnosis that tells them they have a 75% likelihood of cancer, and they should use Drug X to treat it. They need to understand which pieces of information came together to create that prediction or answer. We absolutely believe that going forward we should have complete transparency with regards to ML and think through the ethical implications of the technology advances that will be an integral part of our lives and our society moving forward.
4. Include human beings in the ML/AI learning loop
We need to have human beings in the loop to create the right end-to-end customer experiences. At one point, Redfin, a consumer real estate startup, experimented with sending ML-generated house recommendations to its users. These machine-generated recommendations had a slightly higher engagement rates than a users’ own search and alert filters. However, the real improvement came when Redfin asked its agents to review recommendations before they were sent out. After agents reviewed the recommendations, Redfin was able to use the agents’ modifications as additional training data, and the click-through rate on recommended houses rose significantly. MightyAI, which provides training data as a service for Autonomous Vehicles has been able to deliver very high levels of accuracy and quality data by including a human in the loop.
5. Understand that machine learning is a critical ingredient, but you may not need it on day one
Machine learning is the key ingredient for building today’s intelligent applications, but the first and most important step in building an app is to build a service that resonates with your customers. The advice from people who have done this successfully before is to start with the application and experience that you want to deliver, and in the process, think about how ML can enhance your application based on the standard usage patterns and workflows of your customers.
We are still in the early stages of building intelligent applications, and that makes it an exciting time to be building and investing in startups. Over the next 3-5 years, we hope to see many more startups that create unique user experiences built on machine learning and data, and we are looking forward to meeting those companies!