[ad_1]
Google Gemini, a new multimodal general AI model that the tech giant calls its most powerful yet, is now available to users across the world through Bard, some developer platforms and even the new Google Pixel 8 Pro devices. The new flexible AI model, which comes in three sizes — the yet-to-be-launched Ultra, Pro and Nano — is being seen as Google’s answer to ChatGPT which has been ahead of the game so far when it comes to GenAI.
So, what is Google Gemini?
Demis Hassabis, CEO and Co-Founder of Google DeepMind, says Gemini brings us closer to the vision of “AI that feels less like a smart piece of software and more like something useful and intuitive — an expert helper or assistant”. Gemini has been built from scratch as a collaborative effort by teams across Google. It is also multimodal, which means it is not limited to the type of information it can process and can work understand and operate across text, code, audio, image and video. In contrast, ChatGPT cannot work on video at the moment, at least not natively.
It is also much more powerful than existing models. For instance, Google claims Gemini Ultra’s performance “exceeds current state-of-the-art results on 30 of the 32 widely-used academic benchmarks” used in large language model (LLM) research and development. Gemini Ultra is the first model to outperform human experts on massive multitask language understanding (MMLU), which uses a combination of 57 subjects such as math, physics, history, law, medicine and ethics for testing both world knowledge and problem-solving abilities, it added.
Also, Gemini can “understand, explain and generate high-quality code in the world’s most popular programming languages, like Python, Java, C++ and Go”, the company claims.
Why does Gemini come in three sizes?
Gemini will be available in different sizes to scale it as per the need. Gemini Ultra, the largest and most capable model, will be meant for highly complex tasks. Since this model is still completing trust and safety checks, it is available now only to select customers, developers, partners and safety and responsibility experts for early experimentation and feedback. It will be rolled out to developers and enterprise customers early next year.
Gemini Pro will be best at scaling across a wide range of tasks and is now available in Bard for regular users across the world. On Bard, it has a “specifically tuned version of Gemini Pro in English for more advanced reasoning, planning, understanding and more”. Developers and enterprise customers will be able to access Gemini Pro via the Gemini API in Google AI Studio or Google Cloud Vertex AI.
Gemini Nano will manage on-device tasks and is already available on Pixel 8 Pro, powering new features like Summarise in the Recorder app and Smart Reply via Gboard, starting with WhatsApp. From December 13, Android developers will also be able to build with Gemini Nano via AICore, a new system capability available in Android 14, starting on Pixel 8 Pro devices.
Will Gemini also impact Google search?
Google claimed Gemini will roll out to more products and services like Search, Ads, Chrome and Duet AI. Google said it is already starting to experiment with Gemini in Search, “where it’s making Search Generative Experience (SGE) faster for users, with a 40% reduction in latency in English in the U.S, alongside improvements in quality”.
How does Gemini address the issues of hallucinations and safety?
Eli Collins, VP, Product, Google DeepMind told indianexpress.com that while they have done a lot of work on improving factuality in Gemini, the LLM is still capable of hallucinating. “When we integrate these models with products like Bard, we have additional techniques to improve the accuracy of responses.”
On safety, Google said it is adding “new protections to account for Gemini’s multimodal capabilities” and is considering potential risks and working to test and mitigate them at each state of development. The company claims it has “most comprehensive safety evaluations of any Google AI model to date, including for bias and toxicity” and has conducted research into potential risk areas like cyber-offense, persuasion, and autonomy. It is also working working with a diverse group of external experts and partners to stress-test our models across a range of issues and identify blindspots in Google’s internal evaluation approach
So, is Gemini better than ChatGPT 4?
At the moment it is hard to say, but Gemini seems to be more flexible that GPT4 at the moment. Also it ability to work with video and on devices without Internet give it an edge. Another factor is that Gemini is now free to use while ChatGPT4 is only for paid users.
[ad_2]
