The internet went wild when OpenAI unveiled their GPT-based chatbot ChatGPT, prompting some to declare that Google’s dominance of the search industry is over.
Google vs Open AI
Google, of course, has taken notice of ChatGPT and, as a New York Times investigation had earlier revealed, has already issued a “code red” over the developments. According to the most recent Bloomberg story, the business spent close to $400 million in Anthropic, which is now developing a substitute for OpenAI’s ChatGPT. Similar to the $1 billion that Microsoft put in OpenAI in 2019, this agreement follows one where Microsoft promised to investing almost $10 billion in the firm.
But just how does Google want to launch LaMDA, an AI chatbot that has been in restricted beta testing for a long now? Let’s look at what is known so far regarding Google’s own AI strategy and why it could invest heavily in this in 2023.
Google’s AI acquisition (Google vs Open AI)
Anthropic is not Google’s first significant investment in technology related to artificial intelligence. Google’s parent firm, Alphabet, purchased the British AI research facility DeepMind in 2014. DeepMind is renowned for creating the AlphaGo software that defeated world champion Go player Lee Sedol in 2016 as well as the chess programmes AlphaZero and AlphaFold that accurately predicted the shapes of virtually all proteins known to science.
Other AI businesses that Google has purchased include Alter, Dark Blue Labs, Dialogflow, Granata Decision Systems, Phiar, AIMatter, and Boston Dynamics, according to Crunchbase.
Measured technique that would apply on AI by Google
Google explained its “slower approach” to introducing new AI-based breakthroughs in a thorough blog post it published in January. Additionally, we think that achieving success in AI requires collaboration between ourselves and other parties, such as academics, developers, consumers, governments, regulators, and citizens. If AI is to fulfil its promise for individuals and society, it is imperative that we as a society gain the public’s trust. The statement stated, “As a firm, we embrace the chance to collaborate with others to get AI right.
James Manyika, an SVP at Google, Jeff Dean, the head of Google’s AI division, Demis Hassabis, the CEO and co-founder of Alphabet-owned DeepMind, Marian Croak, the vice president of engineering at Google, and Sundar Pichai, the CEO of Google and Alphabet, all signed it. In addition, the corporation was said to be “doing continual adversarial and similar types of testing and has chosen a distinct and deliberate approach to access and deployment of emerging systems such as LaMDA, PaLM, and Waymo,” according to the report.
Even though ChatGPT is far from flawless, its popularity forced Google management to issue a “code red,” as several employees were concerned that the entrance of this significant technical development may completely disrupt Google’s operations, according to NYT.
Google made the decision to bring in the heavy guns in order to compete with OpenAI’s success with ChatGPT. Company officials met with founders Larry Page and Sergey Brin multiple times to assess Google’s AI product strategy. The urgency that Google is experiencing is highlighted by the firm founders’ first involvement since they quit their day-to-day jobs in 2019.
Because ChatGPT and similar algorithms can provide incorrect, harmful, or biassed information, according to the New York Times, the business has so far been unwilling to share much of their technology. The business sees this as an uphill battle to implement cutting-edge AI without endangering users and society.
In the future, the business intends to make its chatbot’s technology available to other businesses as a cloud computing service. While simultaneously releasing prototypes that fall short of these requirements, it will keep up its trust and safety standards for official goods.
Google AI demonstration (Google vs Open AI)
Google will demonstrate how it is “using the power of AI to reinvent how people search for, discover, and interact with information, making it more natural and intuitive than ever before to get what you need” during an event on February 8th.
A ChatGPT-like chatbot called “Apprentice Bard” is also being developed by Google employees. Because Apprentice Bard is based on the company’s LaMDA technology, it should have certain advantages over ChatGPT. According to CNBC, ChatGPT has little awareness of events that occurred after 2021, but Apprentice Bard incorporates replies from current occurrences.
During the 40-minute presentation, Google’s LaMDA and PaLM language models are anticipated to be on display. And yes, that is the same LaMDA that former Google engineer Blake Lemoine believed to have developed consciousness. The business will make language models like LaMDA accessible for testing in the “coming weeks,” according to Google CEO Sundar Pichai.
In 2023, new Google AI products are anticipated. (Google vs Open AI)
According to the New York Times, Google now intends to introduce more than 20 new products and show off a chatbot-enabled search engine this year. At the company’s conference in May, reviewed prototypes for several of these goods are anticipated to make their debut. This could contain an updated version of the experimental software AI Test Kitchen, which is used to test product prototypes, and Image Generation Studio, which generates and alters images.
Other initiatives being worked on by the firm include the Shopping Try-on feature, a YouTube green-screen function that lets video makers build backdrops, a Pixel smartphone wallpaper generator, and a tool that might sum up videos by producing a new one.
Many of these AI algorithms will likely be made available to software developers and other businesses, which might increase Google Cloud’s income.
LaMDA is used by Google’s Wordcraft project to create fiction from writing submissions. The business is also experimenting with leveraging AI to produce music and films. For instance, the company’s MusicLM can instantaneously produce music based on a text-based suggestion. It’s interesting to note that MusicLM can even read descriptions of photographs to produce music that blends nicely with the scene.
The most intriguing part of Google’s efforts to better integrate AI into its operations will be how it uses the same technology that powers ChatGPT and LaMDA to enhance its search engine. However, many industry analysts think that the corporation will continue to improve the search engine incrementally rather than completely redesign it.