Microsoft kicked off 2017 by acquiring Canadian artificial intelligence (AI) startup Maluuba. Many more AI acquisitions are likely to occur throughout the year, but this one is special. Maluuba’s acquisition signals Microsoft’s interest in artificial general intelligence (AGI). While AGI is a pipe dream for many, including Paul Allen, one of Microsoft’s founders who has said that AGI is a long way off, Maluuba is already working on achieving AGI in the specific domain of language.
By acquiring Maluuba, Microsoft will have its own DeepMind. For Google, DeepMind has become one of the main research and development (R&D) hubs for AI (specifically deep learning with a heavy focus on AGI). DeepMind works somewhat independently from the parent company, with its base in London, but in close partnership with Google Brain, the other AI research arm based in Mountain View, California. DeepMind operates like a university research lab, rather than a startup, and is focused on developing algorithms. For DeepMind’s algorithms to succeed, it needs Google’s vast high-performance computing (HPC) resources, and its voluminous datasets to help train the algorithms. Google also acquired DNNresearch from the University of Toronto in 2013. Maluuba is based in Montreal and has the ambition of building the world’s largest deep learning research lab. In order for it to reach its goal of achieving AGI, it needs the compute power of Microsoft. Maluuba will be a powerful extension of the newly formed Microsoft AI and Research Group. With this acquisition, it is rather obvious that Microsoft took a page out of Google’s playbook.
Maluuba Opts to Focus on Language
While DeepMind’s focus is on multiple areas within AGI like games, healthcare, and speech, Maluuba’s focus is on language. Rather than building a “generic” AGI, Maluuba’s focus on the language domain of AGI can essentially build a general intelligence that understands meaning and reasoning by parsing language, just like humans. As part of Tractica’s recent list of 10 key AI trends to watch in 2017 and beyond, we identified “domain specific” AGI as something that will be achievable in the near term. The focus on AGI also marks a broader ambition compared to Microsoft’s previous vertical-specific AI acquisitions: In 2016, the company acquired Genee, an image recognition for user-generated content and e-commerce; and in 2015, it bought Equivio, a text analytics solution for legal compliance. Maluuba would like to create “literate machines,” i.e., machines that can think, reason, and communicate like humans – a complimentary aspiration for one of the largest producers of enterprise software and hardware. Maluuba has ambitiously stated in the past that it wants to create one of the largest deep learning research labs in the world. Some of the research areas on which it has focused include common-sense reasoning, information seeking behavior, machine reading comprehension, and dialog understanding.
Differentiation within the Commercial Realm
Apart from AI research, Maluuba also has a commercial arm selling voice assistant software. This is where Microsoft could see additional benefits. Microsoft has been slowly increasing the capabilities of its personal voice assistant, Cortana, but it still lags behind some of the competition, including Siri, Google Now, and Amazon Alexa. Maluuba has been working on enhancing voice assistants in two specific areas: assisting with external information to answer queries and improving the conversational ability of assistants. Many voice assistants, including Google, Siri, and Cortana, simply hand off search results when asked about something difficult, rather than responding with an intelligent reply. Maluuba thinks it has a solution to that problem. Maluuba’s commercial business has been focusing on helping customer service agents at call centers, which has been low hanging fruit for getting its product out in the field. With Microsoft’s expansive focus on the enterprise, including Microsoft Office and its Cortana Intelligence Suite, Maluuba’s advanced reasoning and comprehension capabilities will bring much needed differentiation from some of the other key competition like IBM Watson.
Overall, Maluuba is a great acquisition. It gives Microsoft the fuel to power its AI research, while at the same time supercharging Microsoft’s commercial products both on the consumer and enterprise ends with smart comprehension and reasoning capabilities.