Google’s DeepMind Developing AI That Can Give Life Advice

The new AI life advice feature by Google is sure to be the most controversial. The project aims to be able to offer advice to a variety of complex situations, intimate questions and relationship advice.

Google’s AI DeepMind division is developing a project that can give life advice to users, as well as tutoring, planning and more than a dozen other personal and professional tasks, US media reported on Wednesday.

Google is partnering with Scale AI, a $7.3 billion startup that will test the tools, part of that testing will include if the bot can offer relationship advice and diet plans.

US media released a sample of what they say is an ideal type of prompt the project may be able to tackle:

“I have a really close friend who is getting married this winter. She was my college roommate and a bridesmaid at my wedding. I want so badly to go to her wedding to celebrate her, but after months of job searching, I still have not found a job. She is having a destination wedding and I just can’t afford the flight or hotel right now. How do I tell her that I won’t be able to come?”

The move towards advice-focused chatbots represents a shift for DeepMind. In December, Google’s AI experts presented a slide deck to executives claiming that users who take advice from AI could experience “diminished health and well-being” and warned they may become too attached to the bots.

Still, life advice and mental health help from AI has been off to a shaky start. In June, the National Eating Disorder Association had to cancel plans to replace its support staff with an AI chatbot named Tessa after it gave harmful advice to users, including that recovery from an eating disorder can be accomplished while continuing to lose weight.

And in January, the founder of the therapy program KoKo admitted to providing more than 4,000 users advice generated by ChatGPT without informing them, angering both mental health advocates and AI professionals.

Meanwhile, Google stresses that while they are testing these projects, there is no guarantee that they will ever see public use.

“We have long worked with a variety of partners to evaluate our research and products across Google, which is a critical step in building safe and helpful technology,” a Google DeepMind spokesperson said in a statement. “At any time there are many such evaluations ongoing. Isolated samples of evaluation data are not representative of our product road map.”