Representative image
Google’s Gemini has already got enterprises and start-ups excited as they find a worthy alternate to OpenAI’s GPT-4, said industry experts, coming few days after the technology giant unveiled its largest and most advanced AI model to date.
Analysts unanimously called this a big leg up for Google in the ongoing generative AI battle with rivals Microsoft-OpenAI’s GPT and Meta’s Llama 2. The tech giant is now back in the game and could very well be able to take the lead again as it looks to integrate Gemini into its industry-leading offerings like Search, Google Ads, and cloud platform services, among others.
Not to forget, Google’s AI researchers had also pioneered the transformer technology, which was meant to be used for the concept defined as ‘self-attention’ to significantly speed up and augment how computers understood language. Transformers are also the core component of GPT and stand for the ‘T’ in Generative Pre-Trained Transformers (GPT).
Gemini is a multi-modal AI model, which lets it understand and work with different types of information, including text, code, audio, image, and video, simultaneously.
Mahesh Makhija, Partner and Technology Consulting Leader at Ernst & Young (EY) India, told Moneycontrol, “It's just a few days, the initial set of clients I've spoken to are quite excited. There wasn't an alternative to GPT-4 until now. And now there is seemingly a credible option. Many of our clients who are already on Google Cloud will now start to make use of these models.”
According to Chirag Dekate, VP and Analyst, Quantum Technologies, AI Infrastructures and Supercomputing at Gartner, a combination of these (multimodality) capabilities is bound to unlock enterprise use cases that are beyond the (what now seems) legacy text prompts.
“Gartner anticipates that Google will also enable access to Gemini Pro and Gemini Ultra through its Google Vertex AI suite, enabling partners and enterprises to build their powerful applications integrated with Gemini,” he told Moneyconytrol.
Dekate added that the technology ecosystem is on the cusp of a fundamental disruption in the market powered by Gemini that could transform entire industries, from education to software development and beyond, in a way that was not possible using other models.
Story continues below Advertisement
Gemini will be available in three versions: Ultra (for highly complex tasks), Pro (for scaling across a wide range of tasks), and Nano (on-device tasks).
At present, Google’s chatbot Bard is integrated with the Gemini Pro version. Gemini Nano will also be available on Pixel 8 Pro phones through its latest Android update.
The Pro version, meant for developers through the Gemini API in Google AI Studio and Google Cloud Vertex AI, will be live on December 13. Gemini Ultra is currently being tested with select customers, developers, partners, and safety and responsibility experts before being made available early next year.
While Gemini is almost on par with GPT-4, the fact that GPT-5 is at least a year away makes it the first formidable rival to GPT-4, said Sanchit Vir Gogia, founder and Chief Executive Officer (CEO) of Greyhound Research.
“OpenAI will struggle a bit from here on to keep that competitive edge as Google has the resources to quickly get its technology into production and gain scale,” Gogia added.
As per Google DeepMind, Gemini can beat GPT-4 on 30 out of 32 standard performance measures, though the margins are still narrow.
Shift in Google’s business model
Makhija believes that Gemini will also bring some fundamental changes to Google’s business model as it starts embedding the technology into all its products by next year.
Taking a cue from how Gemini can boost its search engine, he explained, “Until now, when you search, you get 10 blue links from Google. And you have to figure out which one is the right one for you. With GenAI, potentially, you will get all of that summarised with an upfront answer. Now, things like these also change a lot in Google’s business model because Search is primarily advertising-based. The entire economics of search works on who's clicking on what link.”
Also read: Gemini vs GPT4: Is Google’s new AI model as good as it claims to be?
Gemini will find extensive use for other branding purposes like storytelling and products, said Faisal Kawoosa, Chief Analyst, Techarc.
“Gemini appears to be good for some complex use cases too. I have always wondered about AI being leveraged to determine the genuineness of products based on an inspection of their physical properties. Gemini will be good for that,” he added.
According to Gogia, integrating Gemini into offerings like sales and go-to-market and adding other features on top of that will be a huge shift, even for the Google Cloud Platform (GCP) and its adoption.
“GCP is already the default cloud service provider for a lot of SaaS customers, enterprise customers, and most AI developers,” Gogia added.
Dekate said, "For the Google Cloud Platform, Gemini will likely enable an unprecedented growth opportunity as enterprise clients seek to disrupt their industries. For enterprises, there is only one question that they need to answer: how soon can they maximise value from multimodal generative AI techniques? If enterprises are waiting for a signal to get involved in Generative AI, this is it.”
Event alert: CNBC TV18 & Moneycontrol will be hosting the most definitive summit on Artificial Intelligence with top global and Indian voices on December 16th in Bengaluru. Click here to register and get access to the 'Palo Alto Networks presents CNBC-TV18 & Moneycontrol Global AI Conclave'.
Google’s ‘multimodal’ AI model Gemini: What it means for enterprises? - Moneycontrol
Read More
No comments:
Post a Comment