Gemini is pulling ahead of ChatGPT – combining Deep Research with Audio Overviews is one of the best uses of AI I’ve seen so far

(Image credit: Shutterstock/mundissima)

Last week Google announced that not only was Deep Research, Gemini’s research tool, being made available for free to all Gemini users, but it was also adding Audio Overview (its AI-generated podcast feature) to Gemini as well.

What’s amazing is how well these two tools combine together: First, you use Deep Research to find out about a complex subject and generate a report on it that breaks everything down and gives you all the answers you need.

Next, you get Gemini to create an Audio Overview, an easy-to-listen-to AI-generated podcast that explains the report to you, so you don’t even need to read it. The result is that you’ve leveled up your understanding of a subject without having to do any of the research work yourself.

I know kung fu

Keanu Reeves is playing Neo in The Matrix holding out his hand as objects float in the air around it. — (Image credit: Warner Bros)

Until they invent neural links we can implant in our brains to download information, Matrix-style, using Audio Overview, combined with Deep Research, is probably the closest we’ll get to downloading information directly into our heads.

Let me give you an example. As a Brazilian Jiu-Jitsu practitioner, I was quite interested in the percentage of BJJ matches that were finished by an armbar (a joint lock) compared to other submissions, so I got Gemini’s Deep Research to create a report on the issue.

It analyzed all the major BJJ tournaments of the last few years from which data was collected and drew some conclusions, complete with sources.

It was pretty impressive, but the report was still quite dense and intimidating, so I uploaded it back into Gemini and clicked on the ‘Generate Audio Overview’ chip that popped up.

It took it a couple of minutes, but it then produced a six-minute podcast that explained all the key points to me. After a quick listen, I had downloaded everything I needed to know about the Matrix.

Pick your subject

You can pick your subject – it doesn’t matter what it is – it could be anything, but the way these two technologies combine together is just perfect for learning, and it’s edging me towards using Gemini more often than other AI chatbots.

ChatGPT is still great to use of course, and it recently upgraded the quality of its voice communication, but thanks to Audio Overview, I think Gemini has the edge on it right now.

In fact, I’ve set up the Action button on my iPhone to instantly fire it up and load up its voice mode automatically. That way I can talk to Gemini at the press (and hold) of a button, rather than having to load up the app manually.

So, if I want the answer to something, I can just ask Gemini directly, and it replies without having to ask me if it would like to pass that request on to ChatGPT, which is Siri's default action these days.

In comparison, Gemini is fast, efficient, and just works.

It's ironic that I'm having to use a Google app to make my iPhone feel complete, but with Apple's Siri upgrade delays meaning that it's unlikely to fix the situation soon, it feels like Gemini is the best option right now, even for iPhone users.