Artificial Intelligence progress has been uneven in the past 5 decades, with spurts of breakthrough followed by “winter-like” periods of stagnation. Barriers like technology cost, organization capability and ineffective policies held back AI from going mainstream.
Yet, this trend shifted starting in 2005 with performance breakthroughs across numerous cognitive applications, largely due to convergent advances across three enablers: computing power, training data, and learning algorithms.
AI Goes Mainstream
So is the AI boom-and-bust era finally over? In our recent whitepaper “Technology and Innovation for the Future of Production,” co-created with The World Economic Forum, we contend that AI solutions will be widely adopted over the next 10 to 15 years—and advances in hardware, software, and social infrastructure will unlock faster, smarter, and more intuitive applications to make value chains more productive. As AI transitions from rigid, rule-based algorithms to flexible, self learning ones, its impact on the nature of consumption and the structure or firms, supply chains, and production will be profound. Here are some key themes:
1. Products and services will compete on personalized, cognitive features. A P&G shampoo app might automatically adjust your anti-dandruff formulation over time to improve its effectiveness. A Samsung refrigerator might customize your next grocery order based on nutritional and budgetary considerations. Or a Medtronic artificial pancreas with a built-in glucose monitor might automatically administer properly dosed insulin to type 1 diabetes patients.
2. Organizations will become more efficient hierarchies. AI applications will enable organizations to reap the benefits of scale without sacrificing agility—for instance, by allowing them to rapidly simulate decisions across silos, spans, and layers. Already today, AI apps can almost instantly scan reams of data in unstructured Excel and PDF formats to identify S&OP outliers and compliance risks.
3. AI-as-a-service platforms will appear. These platforms, using a pay-as-you-go pricing model, will allow firms to scale up cognitive solutions at a marginal cost of zero, which will cause traditional barriers to entry to come tumbling down. Amazon, for example, is offering smaller brands access to its cloud-based web commerce and fulfillment assets, helping them to grow rapidly with no need for upfront capital. As Amazon embeds AI for smarter predictive forecasting and basket recommendations, a symbiotic relationship may develop: smaller vendors will benefit from Amazon’s capabilities, while Amazon will benefit from a wealth of training data.
Proceed with Caution
Before racing to invest in yet another AI start-up or internal incubator, Executives must first understand what AI can and cannot do for their specific domains and issues. Furthermore, AI’s deeper macroeconomic and social implications must also be kept in mind. According to the January 2017 World Economic Forum white paper “Realizing Human Potential in the Fourth Industrial Revolution” concerned stakeholders will need to address the issues of constantly changing skill requirements and emerging work formats, with the attendant need to revamp the entire education, training, and work ecosystem. And of course, the big question-mark and challenge is speed, since AI solutions are both scalable and exponential, coupled with our limited ability to foresee the new jobs of the future.
A Road Map for Businesses and Governments
How can businesses and governments tackle these challenges and ensure that AI will benefit society as a whole? A few of the ideas outlined in the white paper are that leaders should:
· Make the education ecosystem more responsive to emerging needs. In particular, they should develop policies for early childhood and lifelong learning, update curricula, provide early exposure to the workplace (for example, through internships, mentoring, and access to employer networks), and promote digital fluency.
· Adapt how work formats are regulated and social protections are provided. Employment law and regulatory classifications must be reformed to better accommodate independent workers. Workers will need to be enjoy social coverage while they navigate from job to job. And employers will have to develop the capability to engage with talent in new ways, including by tapping into freelance communities and embarking on more collaborative ways of working.
Although it has not yet been proven that the downside risks of AI adoption will outweigh the known benefits, one thing is certain: understanding AI’s potential and risks and reconsidering the educational and work ecosystems in light of the technological revolution is the beginning, not the end of the process of building sustainable companies and societies. In this journey, business, governments, experts, academia, and other concerned stakeholder must adopt a truly collaborative approach to find solutions.
– Article originally posted on my linkedin site here