News

Nvidia’s new software tools will take care of the language of chatbots

There is no doubt that coexistence with chatbots is a reality in the virtual universe. Companies have sought for a long time how to stand out from the competition by offering better service and optimal customer service that make them stand out as a benchmark company in the sector.

Chatbots have become popular among consumers being integrated into websites or applications. They are defined as a virtual assistant that communicates with them through text messages providing them with help. Thanks to artificial intelligence technology, the user can have a conversation in real time through software integrated into a messaging system.

Without a doubt, they are an ally to solve the problems that consumers have in their commercial experiences, but sometimes they can also play a trick on them. As there is no physical person behind, It is difficult to control what type of answers they give at each moment. They can provide help at any time, but is everything so simple?

In response to this we can say no. Chatbots, although they are very helpful, can play tricks since they are totally unpredictable. Sometimes they can give answers or say things that their creators would not wantgiving false information or producing erratic responses, and this is where the new update to Nvidia’s software tools comes in.

Nvidia bets on language care in chatbots

Nvidia’s Vice President of Applied Research Jonathan Cohen stated that “security in generative artificial intelligence is an industry-wide concern.” Faced with this real problem, the world’s leading hardware and software company, which is also a chip provider, presented on Tuesday a set of software tools designed to help chatbots and thus take care of the language used by them.

In this sense, the American company that was founded by the Taiwanese electrical engineer Jensen Huang in 1993, has provided help to companies such as Microsoft, specifically the search engine Bing, to incorporate human-like functions. However, when it comes to chatbots, they are less predictable in their operation.

So much so that in February, Microsoft limited the number of questions per session to its users with its Bing search engine, after the New York Times reported that it gave disturbing answers. Another aspect to take into account is that these virtual assistants can give among their responses mention the top competitors, and this is one of the aspects to take care of.

With all this, Nvidia has developed these software tools to distribute them for free in order to protect companies from these unwanted responses that their chatbots can give, controlling the conversations between users and virtual assistants in terms of security. and protection.

With this, thanks to the upgrade of the chip manufacturing company, efficient security measures will be activated to prevent chatbots from answering dangerous questionssuch as how to create weapons or the spread of computer viruses.

Surveillance in this regard focuses on two basic elements that Nvidia have developed, the theme, ensuring that conversations stay on topic within a consistent tone; and safetyavoiding dangerous or inappropriate language, as well as connecting to unsafe third-party applications.

With this, Vice President Jonathan Cohen, was optimistic with this new launch of his firm, stating that: “Our goal is to allow the ecosystem of large linguistic models to evolve in a safe, efficient and useful way.”

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *