MADRID, 3 Feb. (Portaltic/EP) –
Microsoft has reiterated its commitment to the development of a ethical artificial intelligence (AI) that helps the development of society as a whole, in a context in which the appearance of the OpenAI ‘chatbot’ ChatGPT has revolutionized the achievements and possibilities of this technology.
“The innovations in AI that we expected for 2033 will arrive in 2023,” said the president of Microsoft, Brad Smith, in an editorial related to the moment that this technology is experiencing, for which this year could “mark a critical turning point“.
Microsoft has been collaborating with OpenAI for years to bring the latter’s language models to its solutions. In 2020 it integrated GPT-3 into Azure, then considered “the world’s largest and most advanced language model”, with 175 billion parameters, which was being trained by Azure’s AI supercomputer.
Recently, Microsoft has announced an extension of the agreement by which it will include ChatGPT technology in its Azure OpenAI service. This new OpenAI ‘chatbot’ is based on GPT-3.5 and has the ability to return long, coherent responses that are not easily identified as machine-generated responses.
Smith says that in the early summer of last year they began testing “an OpenAI technology similar to what the world now knows as ChatGPT.” Also Microsoft’s internal AI services. And this experience has led them to verify that “the opportunities it offers are enormous”.
“We need to take advantage of this decisive year not only to release new AI advancesif not then responsibly and effectively address both the promises and the dangers that lie ahead“, says the manager.
Smith believes that “AI may represent the most important technological advance of our lifetime”; he believes it is a “powerful tool for promoting critical thinking and stimulating creative expression” and for seeking and discovering information.
“AI will boost productivity and stimulate economic growth. It will reduce monotony in many jobs and, when used effectively, will help people to be more creative in their work environment and have a positive effect on their lives,” he defends.
However, Smith is aware that this technology has risks, since it can also become a weapon with which to “deliberately attack society with false information, undermine democracy and explore new ways of doing evil.” For this reason, he stressed that at Microsoft they have been working since 2017 on the creation of “a responsible AI infrastructure”, that they develop together with cybersecurity, privacy and digital security.
Facing the future, the continuous advancement of AI, the company has set itself three objectives: the development and responsible and ethical use of this technology, the promotion of international competitiveness and national security, and making AI serve the society broadly, not limitedly.