Samsung NEXT has announced the launch of Samsung NEXT Q Fund to fund startups involved in artificial intelligence (AI). Q Fund will provide Seed and Series A financing to startups solving AI problems, as well as those using AI to solve computer science problems.
Vincent Tang, principal at Samsung NEXT Ventures said, “For the past 10 years, we’ve watched software eat the world. Now, it’s AI’s turn to eat software,” . “We’re launching Q Fund to support the next generation of AI startups [that] look to scratch beyond the surface of what we know today.”
Instead of traditional venture capital (VC) investments in AI, Q Fund will mainly focus on non-obvious, forward-thinking approaches to AI in spaces, including learning in simulation, scene understanding, intuitive physics, program-learning programs, autoML, robot control, human computer interaction and meta learning.
The investments will help Samsung NEXT achieve its main goal: focusing on what the world will look like in five years and beyond, then helping make that a reality.
“There are multiple approaches to building fundamental AI technology,” said Ajay Singh, director at Samsung NEXT Ventures. “And we want to invest in the people and teams who will try new approaches to lay the groundwork for what AI will be. For this reason, Q Fund will prioritize technical diligence over revenue models.”
To showcase its desire to invest in startups that are looking to replace conventional algorithms, data structures, and general purpose computer programs with learned systems, Q Fund has already invested in Covariant.AI. The developer draws on advances in imitation learning and deep-reinforcement learning to teach robots new, complex skills.
Pieter Abbeel, professor at Berkeley Electrical Engineering and Computer Sciences, and founder of Covariant.AI, said, “We’re at a critical juncture in the development of robotics — learning approaches are about to open up a wide range of new applications. Q Fund understands how many ‘grand challenge’ problems, including drastically expanding robotic capabilities, will be solvable with an AI-first approach.”