OpenAI Officially Releases Sora – An AI Video Creation Tool That Shook Hollywood

Welcome, Growth Pioneers! 🚀

Google has just launched a new AI model, Gemini 2.0 Flash Thinking Experimental, focused on multimedia understanding, reasoning, and programming capabilities. They also introduced the Deep Research tool, supporting advanced research for users.

Bytedance is also joining the trend by introducing INFP, a unique AI technology that transforms images into vivid characters that can speak and sing from any audio file, promising to revolutionize the podcasting field.

This Email Newsletter will cover:

• Google launches new “Reasoning” AI model: Gemini 2.0 Flash Thinking

• Deep Research suggests Google may win the AI race

• Truly astonishing! Bytedance launches INFP – AI that enables images to speak and sing from any audio file!

GEMINI
GOOGLE LAUNCHES NEW “REASONING” AI MODEL: GEMINI 2.0 FLASH THINKING

Source: Google

 

Google has just announced a new AI model called Gemini 2.0 Flash Thinking Experimental, currently in the experimental phase on the AI Studio platform.

According to the description, Gemini 2.0 is optimized for multimedia comprehension, reasoning, and programming, with the ability to solve complex problems in fields like programming, mathematics, and physics.

Logan Kilpatrick, head of product at AI Studio, stated that Gemini 2.0 Flash Thinking Experimental is the “first step in the journey of developing reasoning capabilities” at Google. Jeff Dean, chief scientist at Google DeepMind, emphasized that this model is trained to use thinking to reinforce its reasoning abilities.

In initial tests, Gemini 2.0 showed potential but still needs improvement. This model often takes a few extra seconds to minutes to provide solutions due to the deeper reasoning process. For example, when asked about the number of “R”s in the word “strawberry,” the model incorrectly answered “two” instead of “three.”

Gemini 2.0 Flash Thinking Experimental is built upon Google’s latest Gemini 2.0 Flash model and is similar to other reasoning models like OpenAI’s o1. These models can self-check information, helping avoid some common errors in conventional AI models. However, high operating costs and the ability to maintain long-term progress remain challenges for reasoning models.

The launch of Gemini 2.0 has spurred competition in the AI field, with other companies like DeepSeek and Alibaba also quickly introducing similar reasoning models. This trend reflects the need to find new methods to enhance AI efficiency, as traditional scaling techniques no longer yield the same improvements.

DEEP RESEARCH
DEEP RESEARCH SUGGESTS GOOGLE MAY WIN THE AI RACE

Source: Deepresearch

 

Recently, Google introduced a new AI tool called Deep Research, an AI research assistant for Gemini Advanced users. This is a significant step in the development of large language models (LLMs) like Gemini, helping Google catch up to and even surpass competitors in the AI field.

What is Deep Research?

Deep Research acts like a human research assistant, specializing in handling complex and multifaceted questions that require detailed explanations. When asked about the challenges in creating humanoid robots capable of performing untrained tasks, Deep Research not only searches for relevant research articles but also develops a detailed research plan and synthesizes information from various sources to provide a complete answer.

During testing, Deep Research demonstrated the ability to search and synthesize information quickly and accurately, capable of processing large amounts of data thanks to the Gemini 1.5 Pro model with a context window of over 770,000 words. However, there are still instances where the model makes mistakes, such as when counting the number of “R”s in the word “strawberry” and answering incorrectly.

With a combination of powerful resources, extensive experience, and rich data, Google is in a leading position to steer the future of consumer AI. Deep Research is a testament to Google’s strategy of leveraging its internal strengths to develop advanced AI tools.

The launch of Gemini Deep Research has spurred intense competition from other AI companies like DeepSeek and Alibaba, who are also rapidly developing similar reasoning models. This reflects the ongoing need to explore new methods for enhancing AI efficiency and capabilities.

BYTEDANCE
TRULY ASTONISHING! BYTEDANCE LAUNCHES INFP – AI THAT ENABLES IMAGES TO SPEAK AND SING FROM ANY AUDIO FILE!

Source: Bytedance

 

Bytedance just announced INFP, a groundbreaking AI technology capable of bringing any single image to life, expressively speaking and singing from any audio file! This promises to revolutionize how podcasting works.

8 examples of INFP applications:

1. Talking Mona Lisa: Transforming the famous Mona Lisa painting into a conversational character.

Talking Mona Lisa

 

2. Real-time Agent Interaction: Enabling AIs to communicate seamlessly with each other.

3. Interviews: Creating dynamic interviews with character visuals.

Interviews with character visuals

 

4. Singing: AI can transform images into singing artists.

5. Switching between Listener and Speaker: AI can switch roles between speaker and listener.

6. Talking Paintings: Creating paintings that can converse naturally.

7. Enhanced Agent Communication: Expanding the communication capabilities between AIs.

8. Listening: AI can not only speak but also listen and respond.

With INFP, creating dynamic and engaging podcast content is easier than ever. From generating lively conversations between historical figures to presenting authentic and creative interviews, INFP opens up new opportunities for content creators.

PROMPT OF THE DAY
Gather high-quality information from reputable sources for research, writing, or SEO optimization with AI📝

Thank you for listening!

See you next time.

The AI Growth Team 😄 😄 ❤️