Nvidia and Hive Team Up: How They’re Tackling Rogue AI Content Risks
Hive and Nvidia are working together to make AI content safer by combining Hive’s recognition models with Nvidia NIM.
Hive makes custom AI tools that can understand, look for, and create content. Now, in private clouds and on-premises data centers, it is combining its AI models with Nvidia NIM microservices.
As part of the Nvidia AI Enterprise software platform, Nvidia NIM offers models in optimized containers. Its goal is to make it easier and faster to install both custom and pre-trained AI models across clouds, data centers, and workstations.
“Every month, our cloud-based APIs handle billions of requests from customers.” But being able to use our models in private clouds or on-premises has become one of the most popular requests from potential users’ said Kevin Guo, co-founder and CEO of Hive.
Some of the companies that already use Hive are Reddit, Netflix, Walmart, Zynga, and Glassdoor. The first Hive models that can be used with Nvidia NIM are AI-generated content detection models. These models let users spot pictures, videos, and sounds that were created by AI. Hive says that as creative AI tools continue to grow, there is a chance of misrepresentation, false information, and fraud. This makes things harder for insurance companies, financial services companies, news organizations, and others.
As Justin Boitano, vice president of business AI software products at Nvidia, put it, “AI-generated content detection is becoming an important tool for helping insurance and financial services companies find attempts at misrepresentation.” “Businesses can quickly use Hive’s detection models with NIM microservices to help protect their companies from fake content, documents, and claims.”
Also, Hive is letting social networks use its technology for free for 90 days. Guo said that the fact that creative AI tools make it easier to make content can also be dangerous for many businesses and groups. When platforms that use user-generated content try to handle a lot of AI-generated content, they run into different issues. We know how to help you deal with the risks.
Hive is going to add more models to Nvidia NIM “in the coming months.” These will include models for content moderation, logo identification, optical character recognition, speech transcription, and creating custom models through Hive’s AutoML platform.
