Published , by Donovan Erskine
Published , by Donovan Erskine
Artificial Technology has been incredibly polarizing as more of the world’s biggest companies continue to invest money and resources into it. One of the many concerns around AI is its potential spread of false information and its overall threat to security. NVIDIA has long held a pro-AI stance and has now introduced NeMo Guardrails, a new open-source software that targets some of the largest lingering issues with AI services.
NVIDIA announced NeMo Guardrails in a news post on its website today. NVIDIA has broken Guardrails down into three primary functions: Safety, Security, and Topical. Safety Guardrails will prevent AI from providing users with inaccurate information. Most notably, Microsoft’s Bing AI infamously provided the wrong answer to a query in a February demo, creating even more uncertainty around AI technology.
Security Guardrails will “restrict apps to making connections only to external third-party applications known to be safe,” NVIDIA says. Lastly, Topical Guardrails will stop an AI from answering questions that aren’t relevant to its service. The example given is that an AI working in customer service won’t respond to a question about the weather.
NVIDIA CEO Jensen Huang has long been a supporter of artificial intelligence, long before the ChatGPT boom ignited an AI fire under the butts of the world’s biggest companies. In February, he said that the technology was at an “inflection point.” NVIDIA has also used AI in some of its own products and services, such as NVIDIA Broadcast and DLSS.
With NVIDIA’s NeMo Guardrails on the way, it will be interesting to see how the technology is applied to AI services. We’re also curious to see what other AI-adjacent software or products NVIDIA could be introducing down the road. For more on AI and its implementation in the tech and gaming industries, Shacknews has everything you need to know.