Posts tagged "Advice"

For the past several months, Google has been consistently rolling out generative AI tech like the launch of the Search Engine Experience and giving Chrome the ability to summarize news articles. In the future, the tech giant may be taking artificial intelligence into a new frontier: offering life advice.

This information comes from a recently published New York Times article detailing the company’s next big AI project. DeepMind, Google’s own artificial intelligence research lab, has been testing new technology that could effectively turn generative AI “into a personal life coach”. They state that DeepMind is trying to find a way to have their model “perform at least 21 different types of personal and professional tasks”. As we already established, this includes giving people life advice, such as  “ideas, planning instructions, [plus] tutoring tips”. 

For example, you could conceivably ask the chatbot to drum up a workout or meal plan if you want to lose weight. Or maybe you could ask for suggestions related to specific situations like what to do if you can’t make a friend’s wedding.

Intimate knowledge

Technically, these described functions are already present on Google Bard, so this is nothing new. What’s different, according to the piece, is that the company is seeking to improve the AI’s ability to “answer intimate questions about challenges in people’s lives.” To make this whole thing possible, DeepMind apparently brought together “100 experts with doctorates in different fields… [to] assess the tool’s responses”. 

Presumably, this group includes mental health professionals. The problem is that the New York Times doesn't provide concrete details. We don’t have opinions from any of the experts on exactly how good Google’s life coach chatbot is at the moment. 

Still, the main takeaway is whatever the tech giant is cooking up, it appears to be pretty substantial.

An AI arms race

What’s particularly interesting about this life coach project is it flies in the face of company policy. Google’s Bard Privacy Help Hub recommends not to rely on the chatbot’s responses for “medical, legal, financial, or other professional advice.” 

So, why the about face? The New York Times theorizes this could be part of an effort to overcome rivals like OpenAI. As pointed out in the piece, corporations and start-ups have been engaging in an AI arms race ever since ChatGPT launched, vying for the top position in the industry. Google essentially creating a “robo-therapist” may be the edge it needs.

There is, however, a chance this tech will never see the light of day. A DeepMind spokeswoman told the Times that developers are indeed testing the AI, but evaluations are not indicative of a “product road map.”

We will say the idea of an AI that can help with creating robust workout plans or tutoring people in a skill does sound pretty helpful. Hopefully, a breakthrough will come from these experiments.

• TechRadar’s list of the best AI tools of 2023