Science & Technology

Meta Works to Prevent Inappropriate Replies From New AI Chatbots

Facebook-parent has been conducting stress testing of new AI-based chatbots for several thousand hours, but still cannot give an absolute guarantee that the digital thinking system will not say something undesirable.

Meta Works to Prevent Inappropriate Replies From New AI Chatbots

Meta Platforms is currently taking care of responsibility and reliability, actively making efforts aimed at developing and implementing digital products with built-in artificial intelligence into its ecosystem, including, among others, a program trained to formulate thoughts and draw conclusions in the paradigm of the so-called professional worldview of a detective, to have a virtual appearance, similar to the beauty of Paris Hilton, and be a character in armor named Dungeon Master.

Chatbots are an important part of the overall process of consumer interaction with machine intelligence technologies, and at the same time, their existence generates a special and, doubly important, accessible form of communication of the digital thinking system with a person. The mentioned technological giant understands advanced trends and realizes the vector of the development of civilization. AI is not a purely specialized innovation that exists in the space of a separate industry. Machine intelligence in a technological context has the potential to transform the familiar landscape of human activity.

Meta is actively working on the introduction of generative artificial intelligence tools that can process consumers’ information requests and provide, including original materials that meet their needs, in apps used by billions of people every month. Nick Clegg, Meta’s president of global relations, needs to make sure that AI services will be responsible. For the tech giant, it is the issue of responsibility that is particularly sensitive, since for many years the company has faced criticism that its products have a low level of protection.

Nick Clegg said that thousands of hours of work with red teaming – stress testing – were carried out so that chatbots based on machine intelligence had the highest possible level of security. At the same time, he noted that the company cannot provide an absolute guarantee that the digital thinking system will not generate a semantic construct containing inappropriate and undesirable narratives in response to a certain user request. Nick Clegg also stressed that developers continue to retrain.

Meta’s President for Global Relations takes part, among other things, in the initial stages of the company’s work on the implementation of AI, which can be conditionally designated as the phases of the formation of ideas and concepts for the use of advanced technology. In this case, it means the level of activity when top management makes decisions about scenarios for the apply of machine intelligence tools. Nick Clegg explains the possible consequences of launching products in terms of different aspects and dimensions of the external environment. He analyzes the future existence of the AI tool from a legal point of view, considers the moral and ethical side of the issue, and studies the perception of the existence of machine intelligence in the cultural area.

Meta has been using artificial intelligence in its consumer-oriented products for many years. AI takes a special part in monitoring the semantic content of publications posted on virtual platforms that are part of the ecosystem of the technology giant. The company also uses machine intelligence as a tool to combat the spread of information materials that are false in relation to the state of affairs observed in objective reality.

Currently, artificial intelligence is not only a kind of source of opportunities for the technology giant but also a pressure factor that provokes some problems. Now the company should focus on preventing the risks of AI systems functioning beyond the scenario determined by the content of the user request. Also, artificial intelligence in the context of combating disinformation has a kind of dual nature. Advanced technology can recognize a lie, but at the same time, it can generate it and spread it on a significant scale.

Some users have already discovered that Meta stickers created with the help of artificial intelligence and announced together with the mentioned celebrity chatbots can create interesting images. There is no guarantee that the possibility of generating unwanted visual materials by machine intelligence is excluded.

Nick Clegg, a former member of the European Parliament and ex-deputy Prime Minister of the United Kingdom, was hired by Meta in 2018. His task initially was to lead the company’s global lobbying efforts at a time when the tech giant was facing proceedings over potential involvement in election interference and privacy concerns.

Currently, Meta and other brands are developing artificial intelligence systems under the supervision of governments, regulators, and ethics specialists.

The United States adheres to the concept of voluntary agreements with technology companies to protect products based on machine intelligence. At the same time, there is a rapid movement in Europe towards the formation of a regulatory system for the development and use of AI. In June, the European Parliament approved a bill that defines the limits of the use of machine intelligence and introduces fines for violations of the rules of operation of advanced technology.

Nick Clegg says that the use of Meta transparent open-source licenses for some data arrays and large language models on which chatbots are based is the way to ethical scenarios for using AI. According to him, the higher the level of transparency of companies regarding the principles of the new generation technology, the higher the openness about the potential risks and benefits associated with artificial intelligence, controls, and algorithms of actions in this context.

As we have reported earlier, Barclays Analysts Say About Positive Prospects of Meta in AI Industry.

Serhii Mikhailov

2397 Posts 0 Comments

Serhii’s track record of study and work spans six years at the Faculty of Philology and eight years in the media, during which he has developed a deep understanding of various aspects of the industry and honed his writing skills; his areas of expertise include fintech, payments, cryptocurrency, and financial services, and he is constantly keeping a close eye on the latest developments and innovations in these fields, as he believes that they will have a significant impact on the future direction of the economy as a whole.