Google recently unveiled its new artificial intelligence generative model, “Gemini,” marking a significant step in the era of highly capable AI. The CEO, Sundar Pichai, expressed that the world is entering a new era of artificial intelligence with this model.
This announcement comes as Google faces criticism for lagging behind “OpenAI,” which became a leader in generative AI since the launch of its renowned model, “ChatGPT,” about a year ago.
Critics argue that Google’s release of Gemini, accompanied by a substantial promotional campaign, aims to restore its reputation in the AI field, especially after setbacks with its previous model, “Bard,” widely criticized for its dullness rather than artificial intelligence.
Gemini: A Family of Models
Google envisions Gemini not as a single-version model but as a family, catering to various needs and usage scenarios. The family includes “Gemini Nano,” designed for Android devices, “Gemini Pro,” expected to integrate with many of Google’s AI services, and the powerful “Gemini Ultra,” intended for large-scale language processing, competing with OpenAI’s ChatGPT-4.
Gemini, in its current form, primarily handles text-based input and output commands. However, Google executives hope that advanced versions of Gemini will efficiently handle images, videos, and audio in the near future.
Despite Google’s assertion that they’re not merely competing with more advanced AI models, they emphasize their goal of achieving general artificial intelligence, simulating human understanding, comprehension, and perception. According to Demis Hassabis, CEO of DeepMind Technologies at Google, Gemini’s initial models are just the beginning of creating more sentient and aware AI.
Doubts and Criticisms
As with any new Google product, Gemini has sparked doubts and questions about the actual capabilities of the model. The promotional video on YouTube, garnering over 1.6 million views, lacked the necessary transparency. Bloomberg’s Parmy Olson argued that Google misrepresented Gemini’s strength in the promotional video, showcasing impressive abilities that may not be as reliable in reality.
The video demonstrated Gemini’s multimedia capabilities, such as image recognition, game suggestions, and logical questions. However, a disclaimer in the video description mentioned that the time taken and Gemini’s responsiveness were shortened for the sake of brevity, raising concerns about the authenticity of the showcased capabilities.
Manipulation and a Sense of Inadequacy
Critics highlight that while video editing is understandable to some extent, it raises questions about the true potential of the product. Technology companies have a history of manipulating content to exaggerate the effectiveness of their products. Google, by not allowing direct public interaction with Gemini, missed an opportunity for a more transparent and credible demonstration.
Google’s actions suggest that the company is still grappling with a sense of inadequacy in the wake of ChatGPT’s remarkable success. Losing the initial AI race to OpenAI, Google strives to catch up and prove its AI capabilities.
In this age of artificial intelligence competition, despite attempts at misleading tactics, the undeniable truth remains that humanity is on the verge of achieving general artificial intelligence, approaching and even surpassing human capabilities across various domains.
This article is brought to you by DZ Watch.
