Innovate or disappear. The recent progress of Elon Musk’s OpenAI organization is starting to seriously worry Google’s management. Faced with the rise of models based on artificial intelligence, the management of the American firm has just issued a “code red” internally, according to information from the New York Times.
Google soon to be replaced by an AI?
The concerns of the management of Alphabet, parent company of Google, have increased with the opening to the public in early December of ChatGPT. This prototype chatbot can hold a natural conversation with a human, and is able to generate text on just about any topic. An AI capable of thinking at lightning speed and developing complex concepts. Create a diet, set up a profitable business strategy, develop code… ChatGPT is able to meet all the demands of its user.
A major innovation in the field of new technologies, which could overshadow Google. Today, to search for results on the web, the vast majority of internet users turn to Google search. In the future, conversational intelligence may well supplant the online search giant. Requesting a cooking recipe from an assistant will be faster than making a request on the Internet.
“For Google, it’s a bit like the fire alarm has been pulled“, sum up our colleagues from New York Times. Several senior executives are reportedly concerned about a possible technological upheaval in the years to come.
Accelerating AI research
To respond to the acceleration of the development of AI models, Sundar Pichai, the general manager of Google, has held a series of meetings aimed at redefining the firm’s priorities in this area. The work of many groups has been disrupted to respond to the threat posed by ChatGPT, according to a memo and audio recording. In detail, teams from the Research, Trust and Security departments have been reassigned to AI research.
Several employees are now responsible for developing AI models similar to OpenAI’s Dall-E. As a reminder, Dall-E allows you to generate an image with a simple line of text. A disruption of services that comes before a major conference on the subject in May 2023.
Until now, Google was taking its time to implement AI in its services to avoid possible negative biases of the technology. Indeed, erroneous information and, in some cases, hateful or insulting remarks are provided. Until now, the company preferred to perfect the AI before exposing it to its users.
New Google services coming soon?
A robot chat similar to OpenGPT could be marketed in the future to external companies to provide customer service tasks, according to several Alphabet executives during a summit meeting.
Trust and safety standards will be met for official products, but new experiments not subject to these requirements could also be launched. A real AI laboratory inspired by OpenAI which could be offered to 500,000 users. A warning would then be issued, in order to warn Internet users against the potential negative biases of these test tools.