Google’s Gemini AI Model Suggests Company Holds ‘Virtual Monopoly’ on Digital Advertising

BB1kueea

Google CEO Sundar Pichai. AP © AP

Key Takeaways:

Dan Shipper, CEO of Every Media, recently shared a video showcasing his exploration of Google’s Gemini AI models, including the advanced 1.5 version, along with venture capitalist Jesse Beyroutey. Their objective was to leverage Gemini’s capabilities to identify potential investment opportunities in the stock market, providing a glimpse into a future where sophisticated AI models offer extensive insights through contextual prompts.

In their experiment, Shipper and Beyroutey fed Gemini a trove of raw company earnings call transcripts and posed various questions about the businesses involved. The AI model impressively responded with detailed insights, including an analysis of GoGo Inc.’s plans to enhance its in-air Wi-Fi services.

However, Shipper and Beyroutey were primarily interested in receiving stock recommendations. They prodded Gemini further, seeking companies with specific characteristics such as occupying valuable bottlenecks in the value chain and possessing highly scalable business models.

Gemini complied by presenting a list of stocks categorized into Shockproof Stocks, Bottleneck Stocks, and Scalable Stocks. Notably, Google’s parent company, Alphabet, emerged as the top recommendation among Bottleneck Stocks.

According to Gemini, Alphabet’s dominance in the global search market, with over 90% control, virtually monopolizes online advertising—a statement that aligns with Alphabet’s market position but contrasts with its ongoing legal challenges. The Department of Justice has accused Google of illegally maintaining a monopoly in the online search market, a contention that Google likely does not want its AI models to propagate.

When queried about this discrepancy, Google acknowledged that Gemini can occasionally provide inaccurate or misleading information, despite its confident presentation. The company emphasized its commitment to continuous improvement in refining the accuracy of its AI models.

Exit mobile version