*By Carlo Versano* Paperspace wants to make artificial intelligence more accessible, according to its co-founder and CEO Dillon Erb. Erb told Cheddar Tuesday that artificial intelligence is still so new and complex that mainstream companies find the tech difficult to harness. Although A.I. is a "fundamentally transformative technology," he said, "the tools haven't really been built yet to make it accessible for most applications." The Brooklyn-based cloud-computing start-up just closed a $13 million Series A funding round to build up its machine learning toolkits, which it sells to other companies looking to develop their own machine-learning platforms or products ー think A.I.-as-a-service. Erb said he will use a portion of the new financing to market a product offering called "Gradient," which is, essentially, A.I. in a box. Developers and data scientists can use Gradient to run A.I. and deep learning tasks without having to install or build out their own infrastructure. One of the current roadblocks in modern machine learning is that it requires a great deal of processing power, and "actually harnessing that power is incredibly difficult," according to Erb. Gradient helps solve that problem by doing the computing without a dedicated server. Erb said his product will make it easier for smaller, "nimble" companies that want to run machine-learning experiments, but don't have the resources to do it on their own networks. Erb would not reveal any of the clients Paperspace serves, though he said his company touches industries as varied as robotic simulation, cancer research, and media. But most consumers will probably never interface directly with Paperspace's products. In the vein of the early computer coders, Paperspace is building the backbone that may support our interactions with machines in the future. "We're kind of behind the scenes a bit," Erb said. He plans on keeping it that way. For full interview [click here](https://cheddar.com/videos/paperspace-works-toward-accessible-a-i-with-13-million-funding-round).
