Screenshots via the author
Ever since Google released its new version of Bard — the Internet company’s competitor to ChatGPT — with visual capability, I’ve been testing the system out.
One feature I discovered is absolutely mindblowing: from a simple image, Bard can effortlessly recognize, transcribe, and translate ancient writing.
Before, this is a capability that would have required consulting a professor with decades of experience and an archaeology degree. Now, it’s something that anyone can do in seconds from their smartphone.
Here’s how the feature works — and how you can use it the next time you’re traveling, or simply visiting a museum.
Bard’s Recent Change
I can forgive you if you haven’t been following along with every morsel of generative AI news — it feels like breakthroughs come out about five times a day at this point!
Bard is Google’s generative AI chatbot. When it first launched earlier this year, it was a relatively feeble alternative to ChatGPT that briefly tanked its parent company’s stock price.
As with most things Google, though, Bard has continually improved. And Google quietly took a major leap past ChatGPT this month when it made Bard multimodal, allowing it to process images as well as text.
You can try out Bard (it’s free) and this new feature at https://bard.google.com
Translating an Ancient Mosaic
As a pro photographer and AI pro, I immediately started to experiment with this new capability by uploading huge numbers of photos to Bard to see what it could do.
One particular use case blew me away. A family member recently traveled to Israel, and photographed some ancient sites, including this mosaic at Beit She’an.