Updated on 07 December 2023 11:40 PM IST | By Sanjay Thakur
Gemini is an AI model that has been trained to behave in human-like ways, which might promote the debate about the technology’s potential benefits and drawbacks.
Google says that with Gemini’s assistance, its AI-powered chatbot Bard will become more intuitive and adept at planning tasks.
On December 6, Google took the next step in artificial intelligence with the announcement of Project Gemini, an AI model programmed to behave in human-like ways that is certain to fuel debate over the technology’s potential promise and risks.
The deployment will take place in phases, with less complex versions of Gemini known as “Nano” and “Pro” already being used in Google’s AI-powered chatbot Bard and its Pixel 8 Pro smartphone.
Google promises that with Gemini’s help, Bard will become more intuitive and adept at jobs that require planning. Gemini on the Pixel 8 Pro, according to Google, will be able to quickly summarize recordings made on the device and give automatic replies on messaging services, beginning with WhatsApp.
‘Bard Advanced’ will be released in early 2024.
Gemini’s most significant advancements will not occur until early 2024, when its Ultra model will be utilized to create “Bard Advanced,” a beefed-up version of the chatbot that will initially be available only to a test audience.
The AI will initially only be available in English around the world, but Google executives promised reporters at a conference that the technology will eventually be available in other languages.
According to a Gemini demonstration for a group of reporters, Google’s “Bard Advanced” may be capable of unparalleled AI multitasking by detecting and interpreting presentations comprising text, images, and video.
Gemini will also be included into Google’s dominant search engine at some point, though the specific timetable has yet to be determined.
“This is a very important milestone in the evolution of AI, as well as the begginnig of a new era for us at Google.” Demis Hassabis, CEO of Google DeepMind, the AI firm behind Gemini, stated. Google beat out other bidders, including Facebook parent Meta, to acquire London-based DeepMind nearly a decade ago, and has now integrated it with its “Brain” division to focus on Gemini’s growth.
Starting a Debate:
Google touts the technology’s problem-solving abilities as being particularly adept in arithmetic and physics, fueling optimism among AI optimists that it would lead to scientific breakthroughs that will improve human existence.
However, one side of the AI debate is afraid that the technology could eventually replace human intelligence, resulting in the loss of millions of jobs and potentially even more destructive conduct, such is disseminating false information or initiating the use of nuclear weapons.
“We’re approaching this work boldly and responsibly,” stated Google CEO Sundar Pichai in a blog post.
The entry of Gemini is likely to raise the stakes in an AI race that has been heating up over the last year, with San Francisco upstart OpenAI and long-time industry rival Microsoft.
When OpenAI released the free ChatGPT tool late last year, it was already well into building its most powerful AI model, GPT-4, and was backed by Microsoft’s financial muscle and computer capacity. That AI-powered chatbot rose to international prominence, drawing attention to the economic potential of generative AI and placing pressure on Google to remove Bard in response.
OpenAI introduced GPT-4 in March 2023, just as Bard was on the market, and has since been adding new capabilities targeting at consumers and corporate users, including a function unveiled in November that allows the chatbot to analyze photos. It has competed for business against rival AI startups such as Anthropic and even its partner, Microsoft, who has exclusive rights to OpenAI’s technology in exchange for the billions of dollars it has invested in the venture.
So far, the relationship has benefited Microsoft, which has seen its market value rise by more than 50% in 2023, owing to investors’ expectation that AI will become a gold mine for the computer industry. Alphabet, Google’s corporate parent, has also been riding the same wave, with its market value increasing by more than $500 billion, or nearly 45 percent, this year. Despite the recent excitement about Gemini, Alphabet’s stock fell marginally in trade on December 6.
Microsoft’s increased involvement in OpenAI over the last year, paired with the non-profit’s more aggressive commercialization initiatives, has raised concerns that the organization has strayed from its basic goal of preserving humans as technology progresses.
Those concerns were heightened in November 2023, when OpenAI’s board abruptly ousted CEO Sam Altman in a fight over concealed trust difficulties. Following a reaction that threatened to destroy the firm and result in a major migration of AI engineering talent to Microsoft, OpenAI reinstated Altman as CEO and restructured its board.
OpenAI may find itself striving to demonstrate that their technology is still smarter than Google’s with the introduction of Gemini. “I am in awe of what it’s capable of,” said Eli Collins, vice president of product at Google DeepMind.
During a virtual press conference, Google declined to share Gemini’s parameter count—one but not the only measure of a model’s complexity. The most capable version of Gemini outperformed GPT-4 on multiple choice examinations, grade-school math, and other benchmarks, according to a white paper released on December 6, However, there are ongoing hurdles in developing AI models to learn higher-level reasoning skills.
Some computer scientists argue that huge language models have limits since they work by continuously guessing the next word in a sentence and are prone to making up errors known as hallucinations. “With Gemini, we made tremendous progress in what is known as factuality. As a result, Gemini is our most accurate model in this category. “Still there is unsolved scientific problem that are remarked by Collins.