Google has introduced a new Artificial Intelligence (AI) system named Google Gemini. Google formerly had two different AI systems, Google Bard and Google Gemini, but in February 2024 they were combined into one and officially named Gemini. Google Gemini is accessible via a simple Google search, upon which an AI chatbot is displayed. Users sign in with their Google account, enter a prompt, and then engage in conversation with Gemini. However, Google has been criticized for occasionally providing inaccurate information via Gemini.
Google Gemini can also be accessed via a phone that has the AI system, such as the Pixel 8 Pro, or through a website like Gemini Pro, which requires a subscription. Google Gemini is a multimodal AI model, meaning it is capable of processing information from various sources like video, text, images, and audio, and then generates outputs. The AI is capable of complex reasoning and comprehension of nuance, making it more advanced and humane-like than other AI models.
In detail, the Gemini model excels in text output. It is a large language model, running a vast amount of text data while being able to generate new content not present in the original data, like images, audio, and video. The model uses Google’s own AI chips, Tensor Processing Units (TPUs), to perform billions of calculations per second. Despite not being the first AI model to achieve multimodality, generativity, or large-scale language modelling, Gemini is powerful and scalable, with 1.6 trillion parameters—four times more than GPT-4, the previous record holder.
Multiple fields and industries could potentially be revolutionized by Google Gemini, such as education, entertainment, business, and healthcare. Despite these possibilities, Google Gemini raises ethical and social issues like privacy, bias, and responsibility, primarily due to its ability to access and use personal and sensitive data which could lead to identity theft, fraud, deception, and the propagation of biases and errors.
While there is ongoing debate about whether Gemini is superior to GPT-4 in terms of AI capabilities, it is generally considered that Gemini outperforms GPT-4 in tasks of comprehension, question answering, reasoning, and multimedia content processing. However, GPT-4 maintains its strengths in creative text generation and excelling in situations with limited training data.
Under Gemini, the Bard AI can still be accessed by logging into a Google account on the Bard website. Google Gemini is perceived as the next frontier of AI, with vast potential for applications and implications in various fields and industries. However, like most new technologies, there will undoubtedly be controversies and teething issues before it reaches maturity.
