Google's live demonstration of Gemini intensifies the pressure on Apple as AI technology is increasingly adopted by smartphone users.

• Although Google's live demo on Tuesday encountered some bugs, the company demonstrated its leadership in integrating AI technology into smartphones.
• The company's Gemini features for smartphones are real and will be available for testing in the coming weeks.
• Rick Osterloh, Google devices chief, told CNBC's Deirdre Bosa that they have shifted from projecting a vision of where things are headed to actual shipping product.

At the Google Pixel phone launch on Tuesday, David Citron, a product director, demonstrated the AI capabilities of the company's new assistant, Gemini. However, things became uncomfortable when he announced that all the demos were live.

At a gathering of media and analysts at Silicon Valley headquarters, as well as an audience of 100,000 on YouTube, Citron snapped a picture of a concert poster and inquired with his assistant whether he was available the night Sabrina Carpenter was performing in San Francisco.

The demo froze and displayed an error message. Citron attempted to restart it, but the result was the same. Following a brief conversation with the "demo gods" and a phone exchange, the third attempt was successful.

"I found out that Sabrina Carpenter will be in San Francisco on November 9, 2024, but I don't see any events on your calendar for that date."

Google's Gemini features are real and are shipping for testing purposes, showcasing their advantage in artificial intelligence as they become more integrated into smartphone software. While the demo was brief and buggy, it highlighted the potential of AI in the future. Rivals are also preparing consumers for this future, but Google's features are already making strides in this area.

In June, instead of a live demo, presented a pre-recorded video to demonstrate Siri's upcoming advancement in functionality, including the ability to perform actions and comprehend context under its new AI system, Apple Intelligence.

Apple Intelligence is currently being tested by developers, but some of its most significant enhancements, such as image generation, integration with ChatGPT, and key advancements for Siri, have not yet been officially released from Apple's labs.

OpenAI, the company behind ChatGPT, frequently unveils AI advancements but only allows a limited number of individuals to test them.

According to Rick Osterloh, Google devices chief, we have shifted from projecting a vision of future outcomes to actually delivering products, as he stated to CNBC's Deirdre Bosa on Tuesday.

Google's live demos represent a change from the company's previous attempt to showcase Gemini in a demonstration, which resulted in criticism for editing the video.

"Osterloh stated that the items being displayed today would be shipped in the near future and are crucial. Unlike other companies' announcements, these items will be accessible to a large audience."

In June, Apple announced some scripted live testing with media and analysts for Apple Intelligence on current devices. In July, Apple released a preview of some Apple Intelligence functions for developers, including the ability to generate summaries and a new look for Siri that makes the entire iPhone screen glow. However, the preview does not include functions like image generation, ChatGPT integration, and the most anticipated improvements to Siri, which will enable it to perform tasks naturally.

On Tuesday, Google's launch could intensify competition with Apple, as both companies strive to incorporate AI into their operating systems. According to IDC, the number of "Gen AI" smartphones, which have the necessary chips and memory to run AI, is expected to quadruple by 2024, reaching approximately 234 million devices.

Grace Harmon, an analyst at eMarketer, stated in an interview that we learned about Apple's competition today.

As AI technology advances, the market will witness a shift in AI processing with the increasing use of generative AI on phones. Instead of relying on complex models that mimic human output in large data centers, AI features for mobile devices will primarily utilize simpler functions such as summarization and fluency, which will run on the chips already present in the devices.

Google unveiled various functionalities during its 100-minute presentation on Tuesday that are currently unavailable elsewhere.

The example of Citron asking questions about a poster in a photo showcases a technological advancement known as "multimodal AI," which is not a feature that Apple has planned.

The company introduced a feature that allows users to capture screenshots of what they're viewing, and Google will compile that information into searchable notes.

On Tuesday, Google's most significant presentation was Gemini Live, its cutting-edge assistant. During the demo, the technology demonstrated its ability to converse naturally, similar to a human, by adding items to shopping lists or checking Google calendars. According to Osterloh, in the near future, it will be capable of assisting users in conducting in-depth research. Google executives attributed the technology's capabilities to "decades of investment" in AI and its "integrated AI strategy."

Google and Apple both claim that only they can deliver a complete end-to-end experience due to their unique expertise in integrating hardware and software.

Google has criticized Apple's upcoming integration with ChatGPT, stating that it is less private than Google's approach. The company claims that Gemini does not require a hand-off to a third-party AI provider that may not be trusted.

WATCH: Google announces AI rollout with its hardware