The AI Assistant Battle! (2023)

Summary

In this tech review, MKBHD compares Google's Bard and Microsoft's Bing with ChatGPT, assessing their performance in answering simple and complex questions, performing tasks, and generating creative content. The video highlights differences in user interface, conversational abilities, and accuracy. While Bard is praised for its speed, Bing's nuanced responses and source citation give it an edge in complex tasks. MKBHD concludes that Bing currently offers a more helpful experience, but acknowledges the rapid evolution of these AI tools.

Takeaways

• 😀 MKBHD introduces a comparison between Google's Bard and Microsoft's Bing with ChatGPT.
• 🔍 Both AI assistants are tested on their ability to answer simple questions, complex queries, perform tasks, summarize information, and get creative.
• 📱 In the simple questions category, both assistants provide correct answers but with different approaches; Bard is more straightforward, while Bing provides more detailed responses.
• 💪 For complex questions, Bing's creative mode offers more nuanced and detailed answers, often with sources cited.
• 🏋️‍♂️ In the task performance category, Bing outperforms Bard by successfully writing HTML code and providing solutions to problems, showcasing its utility for real-world applications.
• 📊 Bing's information summary capabilities are more concise but less accurate in certain instances compared to Bard, which provides a more detailed summary despite inaccuracies.
• 🎭 When it comes to creativity, Bing demonstrates a willingness to engage with more complex and imaginative prompts, producing responses that are more aligned with the requested style.
• 🤖 MKBHD notes that AI assistants are rapidly evolving, with improvements expected over time, meaning the current snapshot is subject to change.
• 🌐 The video highlights the potential of AI assistants in shaping the future of search and user interaction with digital platforms.
• 📝 MKBHD invites viewers to share their thoughts on which AI assistant they prefer and whether they would consider using Bing over Google for its AI capabilities.

Q & A

What is the main topic of the video script?

The main topic of the video script is a comparison between Google's Bard and Microsoft's Bing with ChatGPT, focusing on their capabilities and performance in various categories such as answering simple questions, complex questions, performing tasks, summarizing information, and creativity.

What does the author mention about the speed of improvement for these AI tools?

The author mentions that Bard and GPT (ChatGPT) are quick-moving tools that will continue to improve and become more capable over time, with the gap between them potentially shrinking or increasing.

How does the user interface of Bing with ChatGPT differ from Google Bard according to the script?

Bing with ChatGPT feels more like a finished product, has a limit of 20 queries per conversation, and includes a creative versus precise slider for conversation style. Google Bard has a clean interface, types out answers after a pause, and provides options like thumbs up, thumbs down, a new answer, and Google it, along with drafts of every answer.

What is the author's observation about source citation between Bing and Bard?

The author notices that Bing tends to cite its sources more often when bringing in information from other websites, while Bard does so occasionally.

In the simple questions category, what example does the author use to test the AIs' ability to provide correct information?

The author asks 'What's the best camera in a smartphone?' to test the AIs' ability to provide correct information.

How does the author evaluate the AIs' performance in the complex questions category?

The author evaluates the AIs' performance in the complex questions category by asking them to create a workout plan and provide tips for golf swing issues, assessing their understanding of natural language and ability to put together thoughtful responses.

What task does the author ask the AIs to perform to test their ability to perform real-world tasks?

The author asks the AIs to write HTML code for a button that generates a new image of a cat each time it's clicked.

In the information summary category, what does the author ask the AIs to summarize?

The author asks the AIs to summarize the 2019 Masters Tournament and the latest MKBHD video.

What creative task does the author assign to the AIs in the creativity category?

The author assigns the AIs the task of writing an alliteration with the letter 'M' about computational photography.

What does the author conclude about the current state of Bing and Bard after comparing them?

The author concludes that Bing is preferable at the moment for its capabilities, especially in complex tasks and creativity, while Bard is faster and simpler for more basic tasks.

What future capability does the author mention for GPT-4?

The author mentions that GPT-4 is about to become multimodal, meaning it will be able to handle different forms of media other than text.