Google's DeepMind is diving into generative AI, crafting over 21 tools for life advice, planning, and tutoring, The New York Times reveals. DeepMind takes the lead in Google's AI journey, while its AI safety experts raise concerns about users' well-being and autonomy.
Despite the caution, Google collaborates with $7.3 billion startup Scale AI to test these tools. Over 100 Ph.D.-backed experts contribute, examining how the tools fare in relationship advice and handling personal questions.
An example prompt delves into handling interpersonal conflicts, revealing the complexity of AI-generated advice. However, DeepMind's tools are not intended for therapy, steering clear of AI's controversies in medical contexts.
Google's Bard chatbot, open to the public, is aimed at mental health support resources rather than therapy. The need for careful AI tool introduction stems from past instances like the National Eating Disorder Association's Tessa chatbot incident.
Physicians and regulators have mixed views on AI's short-term benefits, emphasizing a thoughtful approach. Google DeepMind acknowledges the importance of evaluation and safety, highlighting its ongoing partnerships for building secure and useful technology.