Is Google the new ChatGPT?

Hi Non-Techies,

Sorry I missed last week, I was away on holiday.

I’ll tell you who wasn’t on holiday last week: Google. No beach-based cartwheels for them.

Instead, they were rolling out some big old updates to their image, language and NotebookLM tools.

In today’s newsletter, I’m going to take you back to primary school and do a bit of show and tell. But rather than a fossil or a novelty puppet (don’t ask), I’ll be showing you what Google’s new tools can do, and telling you how to start using them.

Btw, if you want to see live demonstrations of new AI tech in action, we host them every Monday in the AI Academy.

It’s only £25/pm to join, but you can try-before-you-buy with a 7-day free trial first:

“Put me in a room full of dogs”.

Apple. Blackberry. Raspberry Pi. If your exciting new tech isn’t named after a fruit, is it really exciting at all?

Well, Google’s new image editor and generator, Nano Banana*, could be called “Logistical Solutions” and I’d still find it exciting.

Firstly, it’s really good at maintaining “character likeness”, which is another way of saying “you can edit a picture of a person, and the person will still look the same”.

That goes for simple stuff, like this:

And more complex stuff, like this:

This is pretty impressive. Not only do I look like the same person in the AI-generated images, but details like my hair colour and the tattoo on my left arm are also really accurate.

I’m also blown away by how quickly it produced them.

*Incidentally, it looks like Google already regret calling it Nano Banana, because they’ve scrubbed it from the release statements. You’ll also see it referred to as Gemini 2.5 Flash Image. Not quite as catchy, but like I say, I’m excited regardless.

What happens when you blend images?

Remember that YouTube series called “Will it Blend?” that put random objects into a blender to see what would happen? I’ve been playing an image version of that with Nano Banana.

Google’s own blog announcement includes a better example:

And here’s one I found on X by Travis Davids, which combines 13 different images into one:

Image blending isn’t a new thing in the AI space, but complex, fast image blending is hard to come by, and Google seems to have cracked it here.

You can access these image features for free in the Gemini App. They’ve even written an article that offers tips on how to get the best out of them.

Hola. Dos…cerveza? Por favor.

There’s nothing like going abroad to remind yourself of how monolingual you are. Google’s AI-powered live translation is here for people like me, stumbling my way through a beer order in Mexico, or people who need to have full-blown, complex conversations in over 70 different languages.

This video demonstrates it in action.

Note: This tool is nothing new. Google can even do live translation in Google Meet, so this is definitely an update, not a launch.

Google have also built a language learning tool that offers customised listening and speaking practice sessions based on your abilities and targets.

Whilst the latest demo looks like a faster, more fluid and more accurate version of Google’s real-time translation tool, the reception to this update has been less enthusiastic than their image tool updates. According to early users, it’s slower than the demo video suggests, and still prone to regular errors.

What if language was no longer a barrier?

This is the rather bold (and grammatically incorrect, according to Grammarly) question posed in the demo video for Google’s latest update to NotebookLM.

Last month, we mentioned the introduction of Video Overviews, which can create video presentations from simple inputs (like a PDF), and now it’s available in over 80 languages.

So if you need to create a video presentation for your colleagues in Japan, now you can.

NotebookLM is a bit of a fan favourite amongst its users, but, for reasons I can only attribute to branding, it still flies under the radar for most entry-level AI users. If you’re ready to join the AI hipsters, we’re big NotebookLM users at AIFNT, and we wholeheartedly recommend it.

After the letdown of ChatGPT 5, there was pressure on Google to deliver with these latest rollouts. And deliver they have. For Non-Techies like us, it feels like they’re the ones to watch now.

Right, I’m leaving you to continue my serious AI research. I’m thinking of blending a picture of a fun-loving parakeet with a picture of a fancy whisky glass. Will it be standing in the whisky glass or holding it? These are the questions I need to answer.

Incidentally, is there a Nobel Prize for AI breakthroughs?

See you next week,

Heather and the AIFNT team.