By focusing on data, value propositions and people, companies can move beyond experimentation and be ready to take the next step.
After decades of research and development mostly confined to projects in academia and large organizations, artificial intelligence (AI) and machine learning (ML) have found their way into everything from chatbots to tractors, from financial markets to medical research. It has expanded into every corner of the modern enterprise. However, enterprises struggle to move from individual use cases to organization-wide deployments due to inadequate or inadequate data, talent shortages, unclear value propositions, and concerns about risk and liability. I’m here.
This MIT Technology Review Insights report was commissioned by and produced in collaboration with JPMorgan Chase and includes a survey of 300 executives and industry experts from finance, healthcare, academia and technology. Based on interviews with seven family members, this chart summarizes the enablers and barriers to MIT Technology Review Insights. The road to AI/ML adoption.
Key findings of the report include:
Don’t settle for half the story.
Get access to the latest tech news without a paywall.
subscribe now
Already a Subscriber? Login
MIT Technology Review provides an intelligent, independent filter for large amounts of information about technology.
subscribe now
Already a Subscriber? Login
Companies have embraced AI/ML, but struggle to scale it across their organizations. The vast majority of respondents (93%) have some AI/ML projects that are experimental or in use, and may be more widely adopted by larger companies. A majority (82%) said their investment in ML will increase over the next 18 months, with AI and ML closely tied to their revenue goals. But scaling is a big challenge, as is hiring skilled workers, finding the right use cases, and demonstrating value.
A successful implementation requires a strategy of talent and skills. The challenge goes beyond just attracting core data scientists. Enterprises need hybrid and translation talent to guide AI/ML design, testing, and governance, and a workforce strategy that keeps everyone playing a role in technology development. Competitive businesses must offer their workers distinct opportunities, advancements and influences to make a difference. For the wider workforce, upskilling and engagement are key to supporting AI/ML innovation.
A Center of Excellence (CoE) provides a foundation for broad deployment, balancing technology sharing and customized solutions. Companies with mature capabilities (usually large companies) tend to develop their systems in-house. The CoE offers a hub-and-spoke model, leveraging cross-functional core ML consulting to develop broadly deployable solutions alongside bespoke tools. ML teams need to keep up with the rapidly evolving AI/ML data science developments.
AI/ML governance requires robust model operations such as data transparency and provenance, regulatory foresight, and responsible AI. The intersection of multiple automated systems can exacerbate risks such as cybersecurity issues, unlawful discrimination, and macrovolatility to advanced data science tools. Regulators and civil society groups are scrutinizing AI impacts on citizens and governments, paying particular attention to areas of systemic importance. Enterprises need a responsible AI strategy based on complete data provenance, risk assessment, checks and controls. This requires technical interventions such as automatic flagging of failures and risks in AI/ML models, as well as social, cultural and other business reforms.
Download report
This content was created by Insights, the custom content division of MIT Technology Review. It was not written by the editorial staff of MIT Technology Review.
