OpenAI's ChatGPT vs Google's Gemini

A year ago, OpenAI launched its AI-based chatbot, ChatGPT. Today, after several attempts to compete with ChatGPT 3 / 3.5 / 4, Google is launching Gemini — a unique and powerful AI technology it says can surpass ChatGPT-4: deepmind.google/technologies/gemini

Since ChatGPT came out, hundreds of millions of people have tested, used, or subscribed to it in order to immerse themselves in our new world driven by AI models for everything. In the meantime, while many made OpenAI the talk of the town for all things AI, several tech and finance commentators were already criticizing Google for not being the first tech company to release a mainstream generative AI tool.

I would not bury Google when it comes to AI. On the contrary, they are well positioned if you've followed their AI strategic moves. How come? DeepMind! Founded in the UK in 2010, Google acquired it in 2014. For more than 10 years, DeepMind has been a leader in AI studies and the creation of neural network models. Look at their Wikipedia page to get a sense of all their developments and accomplishments.

In November 2022, OpenAI had nothing to lose when launching ChatGPT; meanwhile, regarding AI, Google had to protect its brand, its main revenue stream (Search and Ads), and its position as a trusted tech provider for billions of consumers and millions of businesses.

With Gemini out in the open AI world, Google is putting its cards on the table; hopefully, a real AI competition is finally starting.

What is Google DeepMind's Gemini? It operates on multiple input formats, including text, code, images, and audio. This allows it to perform diverse tasks, such as problem-solving, advice-giving, and answering questions across various fields. Google claims that the largest version of Gemini 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."

Google: "Gemini 1.0’s sophisticated multimodal reasoning capabilities can help make sense of complex written and visual information. Its remarkable ability to extract insights from hundreds of thousands of documents through reading, filtering, and understanding information will help deliver new breakthroughs at digital speeds in many fields from science to finance."

Here's a breakdown of Gemini's strengths according to Google:

• Superior performance: Gemini reportedly outperforms ChatGPT on various academic benchmarks, including reasoning, understanding, and factual accuracy.
• Multimodal capabilities: Gemini can process and generate text, code, and images, making it versatile for various tasks.
• Efficiency and flexibility: Gemini is designed to be highly efficient and integrate seamlessly with existing tools and APIs. It also comes in different sizes and capabilities, catering to diverse needs.
• Safety and responsibility: Google emphasizes ethical considerations in developing Gemini, focusing on factual accuracy, responsible attribution, and safeguards against potential harm.
• Future-proof design: Gemini's architecture allows for future innovation in areas like memory and planning, opening up exciting possibilities.

Here is how Google reports Gemini capabilities vs. ChatGPT 4 in text and multimodal tasks:

The kicker in all this is Google's deployment strategy for Gemini: it will be available in three sizes: Gemini Ultra ("our most capable and largest model for highly complex tasks."), Gemini Pro ("our best model for scaling across a wide range of tasks"), and Gemini Nano ("Our most efficient model for on-device tasks." like Google's Pixel 8 Pro smartphone). That's right, Google Gemini Nano AI is now on smartphones already in the market!

AI needs competition. Google talked a big game today! Let's hope their newly released AI capabilities will, as they claim, bring the heat to OpenAI.