Introduction
Hello, everyone. Welcome to Gui Gu 101. I'm Hong Jun. Every year in May, Google hosts Google I.O. at its headquarters in Shenzhen. This is a face-to-face developer event where Google releases a batch of new products. After the epidemic, I've been attending Google I.O. almost every year. Let's take a look back at the past two years and explore the latest developments at this year's event.
Google's Challenges and Breakthroughs in the AI Landscape
Google's Response to ChatGPT
In 2023, Google officially challenged ChatGPT by releasing a chatbot called Bard. However, Bard made a mistake in its answer when asked about the new discovery of the James Webb Space Telescope, causing Google's market value to drop significantly overnight.
OpenAI's Impact in 2024
On the day of the Google meeting in 2024, OpenAI launched the 4O model, which shocked the world. Many foreign media even suggested that this was a deliberate attack on Google.
Google's Comeback in 2024
After two years, Google made a remarkable comeback at this year's Google I.O. The Gemini 2.5 model was a major highlight, and Google seemed to have broken through and fought a beautiful battle.
Exploring Google's Big Model Gemini
The Soul of Gemini
In this episode, we'll take a closer look at the soul of Google's big model Gemini and see how it can handle the impact of the big model on search. We'll also discuss how Wall Street views Google's position in the stock market.
Guest Introductions
Today's guests are Kemi Kong, the co-founder of Cambio ML, and Sean Way, the founder of HeyReview. Kemi Kong was previously at DeepMind, and Sean Way used to work on Google's voice assistant.
Impressions from Google I.O. 2024
Google's Model Integration
Sean Way was most impressed by Google's integration of the entire model and its product's width and depth. Google covered all the different modalities needed for multi-modal models, from the Gemini 2.5 Pro multi-modal model to the Imogen image production model and the Vue video production model.
Text-to-Video Advancements
Shuang was particularly impressed by the text-to-video feature. Google's VIEW3 model release not only had video images but also sound, which is a relatively high threshold from a multi-modal perspective.
Google's Search Engine Transformation
This year, Google changed the search engine of its entire business model search entry to AI mode. The AI mode combines the search entry with the large model, providing more accurate answers and even the ability to complete tasks like one-click ordering.
AI Search and Its Implications
Testing Google's AI Mode
Sean Way tested Google's AI mode before its release and compared it with Perplexity and OpenAI Search. While Google's AI mode showed improved context understanding, OpenAI Search performed better in his tests at that time.
The Future of Google's Search Business
AI search is a crucial area for Google as its advertising income is the most stable. Google's AI mode has the potential to change its traditional advertising and marketing modes, and it remains to be seen how it will generate new business revenue.
Google's Ability and Will
Kemi Kong believes that Google has the ability to be the best AI searcher among tech companies, but it also needs to be willing to take risks and integrate its products more effectively.
Google's Shopping Demonstration and Business Model
The Shopping Scenario
Google demonstrated a shopping scenario where a user could search for a dress, compare prices, and even try it on virtually. The entire process was streamlined, from searching to one-click ordering.
Challenges and Opportunities
The new shopping model raises questions about user privacy, payment systems, and how Google will charge for its services. It also has implications for SEO and the future of e-commerce.
Google's Advantages in the AI Race
Data and Infrastructure
Google's biggest advantage is its vast amount of data, including user information and website indexing. It also has a strong infrastructure, with its own data centers and TPU ecosystem.
Talent and Innovation
Google has a large pool of talented employees and a long history of innovation. The return of founders like Sergey Brin has also brought back a sense of urgency and focus.
The Future of Google's AI Ecosystem
Product Strategy
Google's product strategy seems to be centered around building an ecosystem around the Gemini model. By launching multiple products at once, Google is exploring different opportunities and waiting to see which ones will take off.
Competitiveness of AI Giants
In the highly competitive AI landscape, the competition on the model level is intense. To evaluate the competitiveness of AI giants, we need to consider factors such as technology, data, talent, and product strategy.
Conclusion
In conclusion, Google I.O. 2024 showcased Google's significant advancements in AI. The company's focus on the Gemini model, search engine transformation, and product ecosystem building positions it as a strong player in the AI race. However, there are still challenges to overcome, such as integrating products, generating new revenue, and maintaining a competitive edge. In the next episode, we'll discuss how Wall Street views Google's position among the AI giants. Stay tuned!
{# 
{#