Sure, advances in cognitive technologies offer exciting prospects for blue-sky innovations. But in the near term, companies may soon have more opportunities to apply artificial intelligence to help solve current business problems.
Register for our course on cognitive technologies
Futuristic artificial-intelligence scenarios may make for compelling reading. But heavy investment in startup companies that are developing or applying cognitive technologies suggests that established industries may find the biggest near-term opportunities in applying these technologies to traditional business issues.
The trend of relentlessly increasing performance in cognitive technologies has continued in recent months, with notable gains in face and image recognition, which can now outperform humans on some benchmarks, and speech recognition, with Google cutting errors by two-thirds.1 As the performance of cognitive technologies such as computer vision, natural language processing, and speech recognition improves, the number and scope of applications continue to expand. For instance, IBM found its forecasts of the availability of sunlight and wind for energy production developed using machine learning to be up to 30 percent more accurate than those created using conventional approaches.2 A machine learning-based analysis of brain images was shown to be 98 percent accurate in diagnosing Parkinson’s disease, compared with 91 percent for conventional visual-analysis methods.3 In just the last few months, researchers revealed projects in which neural networks can predict phenomena as diverse as bank failures,4 rodent-borne disease outbreaks,5 and clinical responses to anti-cancer drugs.6 These are just a few examples of the many valuable applications of cognitive technologies being created to address today’s opportunities and challenges.
A principal way that cognitive technologies can create value for companies is by intelligently automating tasks and surfacing insights that augment human decision making.
Continuing improvements in the power of cognitive technologies may make possible new classes of products, from intelligent virtual assistants to home robots, and will likely spawn entirely new industries. But a novel analysis of recent venture capital investments suggests a large, nearer-term opportunity: serving existing industries with products enhanced by cognitive technologies.
We can sense where the biggest opportunities may lie by analyzing venture capital investments in startup companies that are developing or applying cognitive technologies. Our analysis is based on a schema created at the venture capital firm Bloomberg Beta;7 it divides cognitive technology startups into five broad categories based on the nature of their products:
Core technologies: broadly useful cognitive technology products such as machine learning platforms, natural language processing tools, and computer vision systems
Supporting technologies: hardware or software platforms supporting the development of applications of cognitive technologies
Rethinking humans: new human-computer interface tools such as augmented reality, gestural computing, and affective computing (recognizing or simulating emotions)
Rethinking enterprise: applications for typical enterprise functions such as sales, marketing, security, human resources, and intelligence/analytics
Rethinking industries: applications and tools designed for specific industry sectors such as retail, education, and health care
The biggest funding category by far is those companies building applications for traditional enterprise functions such as marketing and sales. Startups like these have raised nearly $2.5 billion since 2011 (see figure 1), suggesting that the biggest near-term opportunity for cognitive technologies is in using them to enhance current business practices.
Indeed, startups are using cognitive technologies to develop valuable features and capabilities such as intelligent automation, ease of use, and insightful analytics that are superior to what can readily be achieved with conventional information technologies. In the rethinking enterprise category, for instance, marketing-focused startups have used machine learning to improve customer targeting8 and website personalization,9 natural language processing to understand what consumers are saying about television content on social media,10 and speech recognition to gauge the quality of inbound telephone leads.11 The top segments in the enterprise category are marketing, intelligence (including analytics solutions), security and authentication,and sales.12
Companies developing applications tailored for specific sectors such as retail, advertising (“adtech”), education, medical/diagnostics,and media have also received major investments—over $2 billion during the same period.13 In the rethinking industry category, startups providing solutions aimed at the medical and diagnostics sectors are using natural language processing to automate the coding of medical charts for insurance reimbursement,14 machine learning to power mobile care-management apps that tailor their content to better engage patients in their care regimen,15 and computer vision and machine learning to power a simple, low-cost ultrasound device that can automatically diagnose disorders.16 Top segments in this category include retail, adtech, medical and diagnostics, and education.17
This analysis suggests that the applications of cognitive technologies are broad; they can often resemble traditional enterprise applications—with advanced capabilities and performance—rather than specialized cognitive computing products. A principal way that cognitive technologies can create value for companies is by intelligently automating tasks and surfacing insights that augment human decision making. The opportunities for this are huge, spanning all sectors and business functions.
Cognitive technologies present special opportunities in the technology sector itself. These include enabling customers’ cognitive computing initiatives with platforms and tools to make it easier to build and deploy applications. Earlier this year, for instance, both Amazon and Microsoft introduced commercial cloud-based machine learning platforms.18 Intel has released chips tuned for machine learning that vendors such as Dell, Hewlett-Packard, and Lenovo are incorporating into servers.19 And Qualcomm has released processors aimed at deploying trained neural networks in mobile devices, from smartphones to consumer-oriented drones.20 These developments highlight the potential for technology suppliers to meet growing demand for tools and platforms to support the development of cognitive technology applications.
Vendors of enterprise software applications should consider how cognitive technologies can enhance their products. Startups may offer models of how to employ these technologies to make products easier to use, to automate functions intelligently, and to generate greater insight from data.
Corporate IT groups may want to build awareness of and skills in cognitive technologies such as machine learning and natural language processing. They can also begin to assess how to employ cognitive technologies to enhance existing corporate applications to provide greater usability and more valuable insights to users.
Buyers of enterprise software may find it worthwhile to ask their vendors to explain how they plan to take advantage of cognitive technologies to enhance their products’ performance and utility.
Cognitive technologies continue to improve and will likely give rise to entirely new product categories and even industries. In the near term, though, established industries can find significant opportunities in applying these technologies to conventional business functions, as the venture capital investment data suggests.