Google had teased Gemini amidst its other AI platforms, and this AI model of Google has an ascendancy over the existing AI models like ChatGPT, and the firm’s own Bard.
Since the commence of OpenAI’s ChatGPT, Google urged the team to brainstorm the AI models to reality, that led to the roll out of several AI models during its I/O event 2023. Google then incorporated the Artificial General Intelligence (AGI) into its search-engine, Maps, Gmail and the firm is looking to blend the AI into every of its applications and service.
Among the other AI models, Google’s Gemini has been talked about a little bit, but a recent report by Google DeepMind’s CEO ‘Demis Hassabis’ entitles the specific advantage of Gemini over others, even GPT-4.
Google Gemini is one of the large language models of the firm, likely as PaLM-2, that’s generative and has been trained with “reinforcement learning”, that aids the AI to be worthwhile in planning or in problem-solving tasks. Gemini is still in development by the firm’s DeepMind, and is equipped with DeepMind’s AlphaGo, the AI program that made history by defeating a world champion “Lee Changho” in the board game ‘Go’.
Researchers and scientists even were baffled to see a computer defeating a champion, as they thought it’s non-viable at present. AlphaGo did that in 2016 itself, two years after Google acquired DeepMind.
AlphaGo was pioneering at its time, and the model was built upon a form of machine learning called ‘reinforcement learning’, which deals about powering the software the ability to learn and solve tough problems by repeated attempts and feedback on its performance on a particular matter like video games or ‘Go’.
In the game, it used a method called ‘tree search’ to explore and remember possible moves on the board.
Now Google upgrades this AlphaGo’s capability to perform real-time tasks on internet and for end-users, merging it with Gemini. The reinforcement learning of the Gemini allows it to reach maximum capabilities in problem solving tasks, with ability to think beyond the trained data. It’s like AI getting the brain of managers.
It’s still a long road to have Gemini onboard, but it had garnered global attention, due to its expected potentialities. Its anticipated that Gemini might upscale the AI industry, which could bring generative AI’s market worth to $105 billion (₹8.6 lakh crores).
Google and its AI trajectory
Google has been foresightful about the AI for the past two decades, but cautious on its action that prevented unleashing the tech prior to ChatGPT. The launch of ChatGPT got Google to dash into the market with several AI tools and programs, ultimately to safeguard its dominance in the industry.
The firm has recently merged DeepMind with its primary AI lab ‘Brain’, to jointly work through for AI products, tools and services. Nearly “80 or 90 % of the innovations come from one or the other,” says Hassabis, CEO of DeepMind.
Concerns over AI
As people into the tech might know that AI is not yet reliable and might put false information in sanguine way, the same goes for Gemini. Several world leaders, including Elon Musk and other AI experts called for a pause on further developmental of AI or more powerful algorithms to avoid creating something dangerous, that could someday pose a risk comparable to nuclear war or a pandemic.
DeepMind has been exploring the potential risks of AI since before ChatGPT appeared, and Shane Legg, one of the company’s cofounders, has led an “AI safety” group for years.
However, Hassabis says the pause in further development is impractical to enforce, adding “If done correctly, it will be the most beneficial technology for humanity ever,” he says of AI.
But still, users need to be aware and a proper validation on the AI generated content is required.
Kindly add ‘The Inner Detail’ to your Google News Feed by following us!