In a recent interview with the Financial Times, Google DeepMind CEO Demis Hassabis highlighted the rapid trajectory of Google’s AI models, noting that Gemini 3 is currently topping industry leaderboards and serving over 650 million monthly users. While acknowledging the „ferocious“ competition from rivals like Anthropic, Hassabis expressed significant optimism regarding multimodal capabilities and the potential for AI-powered smart glasses to become the industry’s next „killer app“. However, he issued a cautionary note regarding the current investment landscape, describing multi-billion dollar seed rounds for startups without established products as „bubble-like“ and potentially unsustainable, even as he maintained that AI remains the most transformative technology ever invented.
Critically, Hassabis’s reflection on the „talent war“ – where some researchers are reportedly offered $100 million packages—reveals a deepening divide between commercial hype and scientific mission. While he argues that top-tier talent is driven by „unequivocal goods“ like AI for medicine and science, the sheer scale of capital infusion into the sector suggests that the „engine room“ of AI development is increasingly tethered to speculative financial cycles. Furthermore, his estimation that Western companies lead Chinese rivals by only „a matter of months“ underscores a precarious geopolitical race where the pursuit of Artificial General Intelligence (AGI) by 2030 may prioritize speed over the very safety guardrails and „responsible deployment“ he advocates for.
Watch the full interview on the Financial Times YouTube channel.