Thus, Microsoft created a surprise at the beginning of February by integrating ChatGPT into its Bing search engine, in the form of a chatbot capable of answering users’ questions but also generating various scripts on demand. The US IT conglomerate has already invested $1 billion in OpenAI, the startup behind language models that support ChatGPT, and this year signed a new multi-billion dollar agreement with this Californian company at the forefront of technology. Generative artificial intelligence. New Bing technologies, in the testing phase, should be integrated into Windows 11 (the latest version of its operating system on PCs), the office suite (into a program such as Word or Teams) and its Edge browser. During the first week, journalists and experts reported conversations where the computer program became angry with users (or seemed to fall in love). Microsoft has since added safeguards, without slowing down the rollout.
Google surprised by introducing Bard, its AI chatbot, the day before the launch of the new Bing. Bard relies on a scaled down version of its own language model, “LaMDA”, to facilitate testing and to ensure that Bard’s answers are of a “high level of quality”. The tech giant plans to add similar functionality to its search engine, but has been vague about the practicalities. YouTube’s new head, Neil Mohan, says that generative AI will soon be introduced to creators to “enrich storytelling,” but the video platform will take its time to “evolve these tools with well-thought-out safeguards.”
Like Google, Meta has already been working on generative AI, but it doesn’t look like the company is ready to add those technologies to Facebook or Instagram. Mark Zuckerberg, head of the group in California, hopes it will soon make it possible to create “videos, avatars, and holograms” for different platforms. On February 27, he announced the formation of a team to accelerate Mita’s efforts in the field, emphasizing that there was a lot of work to build the “foundations”. He also presented a language model called LLaMA, which will be made available to researchers for further development.
In the meantime, Snapchat will add a chatbot built on the latest version of the ChatGPT API (Programming Interface). Available to subscribers of the paid version, the “MyAI” tab allows users to ask for advice or write messages. But they won’t be able to get him to write their course papers or create inappropriate content.
In the e-commerce sector, Shopify is also using the ChatGPT API for its new chat bot on the “Shop” mobile app. And followers of Instacart, which allows you to shop and deliver, will soon be able to request recommendations from the app for recipes that will turn directly into shopping lists, again thanks to Open AI’s API.
China’s Baidu Group “should be able to complete internal tests” of its rival ChatGPT in March, a spokesperson told AFP in early February. Called the “Ernie Bot,” it will be made available to the “general public” at an unspecified date. They should serve different purposes, from online search to self-driving. Soon after, online sales platform Alibaba announced that it was working on its own generative AI software.
Finally, Elon Musk, the head of Tesla, SpaceX and Twitter, who invested in OpenAI in its infancy, will, according to The Information, consider developing its own competitor to ChatGPT, but without guarantees. The billionaire, who bought Twitter with the idea of freeing expression there, believes that ChatGPT restricts freedom of expression too much in favor of the views of the American left “wake up,” according to the specialized site. So he would have asked the researchers in recent weeks to set up a competing lab.