Alphabet-C CEO: Is "Search is King" no longer the future? Alphabet-C is ready
In the face of the joint siege of social media, short videos, and generative AI, which are all traffic black holes, Alphabet-C, the former internet giant of the search era, has unveiled its "killer move".
Facing the joint siege of social media, short videos, and generative AI, Alphabet-C, the former internet giant of the search era, has unveiled its "killer move".
On February 8th, Alphabet-C announced that its AI chatbot, Bard, will be officially renamed Gemini. At the same time, they introduced a new subscription plan, where users can access their "most powerful model" Gemini Ultra 1.0 by paying $19.99 per month, directly competing with OpenAI's GPT-4.
Alphabet-C refers to Gemini as the "most powerful model ever". In evaluations, Gemini has lived up to expectations and is on par with GPT-4, becoming the only large-scale model that can rival OpenAI's most advanced model.
Alphabet-C CEO, "Chopsticks Brother" (Pichai), recently revealed his thoughts on Gemini and its potential development path in an interview with Wired.
Innate Multimodality
"Multimodality" is one of the favorite features of Chopsticks Brother's Gemini AI model. Alphabet-C claims that this is one of the elements that sets it apart from OpenAI's ChatGPT and Microsoft's Copilot AI assistant (also powered by OpenAI technology).
This means that Gemini is trained on various formats of data, not just text, but also images, audio, and code. As a result, the completed model can fluently use these modes and interact through text or voice responses, or by taking and sharing photos.
Chopsticks Brother excitedly said, "This is how human thinking works, constantly seeking things and truly desiring to connect with the world you see." He stated that he has long sought to incorporate this ability into Alphabet-C's technology. "That's why we introduced multiple searches in Alphabet-C search, that's why we created Google Lens (for visual search). Gemini is inherently multimodal, you can input images and start asking questions. That's what sets it apart."
Free or Paid?
Alphabet-C is also conducting parallel experiments on reshaping its core search interface with AI, introducing a generative search experience that provides chatbot-like answers before the familiar ads and link lists.
Chopsticks Brother stated that Alphabet-C is currently committed to improving the generative AI experience and remains open to both the "free ad" and "paid subscription" business models. He refused to disclose whether the paid Gemini product will be completely ad-free but used their other product, YouTube, as an example. YouTube started experimenting with paid, ad-free subscription services a few years ago. "Ads allow us to reach more people with our products, but there are cases where subscriptions provide a different experience," he added. "I can imagine the same user switching back and forth between free search and Gemini subscriptions." In other words, generative search is no longer an add-on to search, but a main menu item - albeit potentially more expensive.
Not an "illusion," but "imagination"
However, Chopping Brother admits that even with the advanced version of Alphabet-C Gemini, there is still a risk of "illusion" like what Bard or other generative AI applications do. "We want people to be aware of this," Chopping Brother said. "I think this technology is useful for many people. But it must be used in the right way, and I still have concerns about people relying on it."
Chopping Brother said that Alphabet-C is trying to reduce the phenomenon of model runaway. But he also cautioned that the word "illusion" should be used with caution, implying that illusion is both a feature and a bug, which is an interesting rebranding of misinformation. He believes that this technology should be based on facts, but if you tune it too low, your chatbot will become very boring and quickly lose interest.
Chopping Brother said that the generative AI experience should be "imaginative." "Just like a child who doesn't know what limits are when imagining something."