The immediate benefit of RPA is that it can eliminate a lot of repetitive manual labor and free up humans for what they do best. But RPA also helps enterprises create a standardize framework for capturing data about how they execute processes, as well as data about how processes can get delayed or stalled. "If you set up RPA the right way by instrumenting the process, it's possible to gather data to use as the training set for machine learning," said Catalytic chief revenue officer Ted Shelton in an interview at Transform 2019. "RPA is the gateway drug for AI." An RPA implementation not only puts the steps involved in a process into a bot script, it can also set up the framework for understanding how a process is affected by different variables.

One could also argue that RPA lays the groundwork for machine learning and more intelligent applications. It both gathers useful data and is being combined with AI capabilities. One of us (O'Dell) recently interviewed Eric Siegel, a predictive analytics expert and author of the book, Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die. Siegel pointed out an often overlooked benefit of starting by digitizing processes with simple RPA: the digital bread crumbs it now leaves behind. "This data wasn't amassed in order to do machine learning. The transactional residue accumulates and, lo and behold, it turns out this stuff is really valuable because you can learn from it. You can derive these patterns to help improve the very transactional processes that have been accumulating the data in the first place."

Google has designed a low-power version of its homegrown AI math accelerator, dubbed it the Edge TPU, and promised to ship it to developers by October. Announced at Google Next 2018 today, the ASIC is a cutdown edition of its Tensor Processing Unit (TPU) family of in-house-designed coprocessors. TPUs are used internally at Google to power its machine-learning-based services, or are rentable via its public cloud. These chips are specific designed for and used to train neural networks and perform inference. Now the web giant has developed a cut-down inference-only version suitable for running in Internet-of-Things gateways.

We all interact with artificial intelligence every day: It builds our Google searches, our Facebook feeds, and predicts the next word we'll type. But many of us use Excel. In an attempt to make it easier to work with machine learning on a daily basis, which Microsoft has repeatedly claimed will improve our lives rather than kill jobs, the company is adding a slew of machine learning tools to its Excel spreadsheet software, according to TechCrunch. Spreadsheet jockeys will be able to import machine learning models to analyze data within Excel, and the program will automatically recognize items such as company names and locations, and pull in additional data. The models could predict future sales numbers given different scenarios, or stand in for any number of software-as-a-service analytics tools that have become popular in sales and marketing.