Salesforce CEO Marc Benioff has commended Google’s voice-based AI assistant, Gemini Live, labeling it the “future of consumer AI” and praising its swift, zero-latency response times in a social media post.
His positive remarks are in stark contrast to his criticism of Microsoft’s Copilot AI agents, which he derided as “Clippy 2.0,” evoking the memory of Microsoft’s much-maligned animated Office assistant from the late 1990s.
The Rise and Features of Gemini Live
Gemini Live debuted during the Pixel 9 event in August 2024 and then rolled out to Gemini Advanced users. By September, the tool had expanded to most Android devices, enabling broader access to natural, real-time voice interactions.
Unlike traditional AI tools, Gemini Live offers hands-free, spoken conversations and supports multitasking, even when a device is locked. The service’s release on the Apple App Store, in conjunction with the iPhone 16 launch, positions Gemini Live as a direct competitor to Apple’s Siri.
Google’s voice assistant is designed to handle complex, uninterrupted dialogues. However, while praised for its seamless performance, some users have noted that its voice customization options are limited.
Just downloaded Gemini Live, and I’m absolutely blown away. Speaking directly to AI with zero latency—this voice interaction is the future of consumer AI. Incredible work, @sundarpichai. Truly groundbreaking. Try it immediately: https://t.co/PysnzJsh7f”
— Marc Benioff (@Benioff) November 17, 2024
Challenges and User Feedback
Despite its strengths, Gemini Live is not without issues. User feedback highlights that Gemini Live lacks features like variable pitch and personalized inflection, making it sound mechanical compared to competitors like OpenAI’s Advanced Voice Mode in ChatGPT.
Reports have also pointed out the tool’s tendency to generate “hallucinations,” instances where the AI provides inaccurate or outdated information confidently. For example, in practical use, Gemini Live has recommended closed businesses as current options, underscoring the problem of misinformation common to many AI tools.
|
Related: |
Benioff’s Perspective on AI Tools
Benioff’s praise of Gemini Live comes alongside his well-publicized critique of Microsoft’s Copilot, a tool integrated across Microsoft 365 applications like Teams and Word.
Although Microsoft´s Copilots support voice interaction, its primary design is more focused on text-based assistance and productivity tasks, which contrasts with the fully voice-centric experience of Gemini Live and similar tools. Benioff’s “Clippy 2.0” comparison underscores his belief that Copilot does not meet the modern standards of AI engagement.
While large corporations such as Vodafone and Cognizant have embraced Copilot for its utility in streamlining tasks and boosting productivity—Vodafone, for instance, reported that Copilot saved up to three hours per employee per week—Benioff remains skeptical.
He believes that the tool falls short in providing substantial, user-focused value in a way that voice assistants like Gemini Live aim to achieve.
AI Voice Assistants from OpenAI, Apple, Amazon
In the broader landscape of AI voice assistants, OpenAI has made considerable strides. Its Advanced Voice Mode, which expanded to Europe in October 2024, allows users to have near-instant, spoken conversations with customizable voice settings.
This mode responds in about 320 milliseconds, making it one of the fastest available. OpenAI’s tool also supports intonational variations, giving users more control over how the AI speaks—a feature Google has yet to fully implement with Gemini Live.
Apple, meanwhile, has struggled to keep Siri competitive. Leaked internal documents revealed that Siri’s response accuracy was about 25% lower than ChatGPT’s, showing how far behind it lags in the AI (voice) assistant race. These limitations come even as Apple leverages advanced hardware like the new iPad mini to bolster Siri’s capabilities.
Amazon is also looking to capture more of the AI voice assistant market with its upcoming release of Alexa Plus, a subscription-based enhancement set to launch soon. Alexa Plus is expected to offer more tailored, interactive conversations, including moderated features for younger users. This move follows Amazon’s recognition of the increasing importance of voice customization and interaction, aspects that have become central to AI development.
Ongoing Developments in AI Voice Technology
As competition grows, tech leaders are investing heavily in refining voice technologies. Google’s plans to add voice-based document management to Gemini Live, hinted at in recent beta code findings, could set it apart by expanding its utility in multitasking scenarios.
These advancements reflect a broader trend where voice assistants are evolving from simple query responders to multifaceted, adaptive tools capable of managing various tasks through spoken commands. The focus remains on improving personalization, accuracy, and response speed to better cater to user needs.
|
Related: |
AI Voice Assistants vs. AI Agents
While Gemini Live, Siri, and Alexa Plus focus on voice interaction as their primary mode of user engagement, tools like Microsoft’s Copilot and the recently rolled out Salesforce Agentforce service attack the completely different market of AI agents and serve different purposes.
Copilot functions as a productivity booster within Office applications, facilitating tasks like generating text, extracting and combing information or summarizing meetings, while Agentforce automates enterprise-level workflows through data-driven actions without direct voice input.
OpenAI is preparing to debut its own first AI agent, Operator, in January 2025, while Google is set to release “Jarvis”, focusing on browser tasks and integration in Google Chrome. Like Google, OpenAI continues to push for overall improved AI performance through its reinforcement learning approaches, seen in its upcoming and recently leaked o1 model.
|
Related: |
