ChatGPT’s Evolution: From Text to Vision
For some time now, we’ve been discussing in these articles how AI is set to become a significant support tool for all professionals, especially analysts.
Well, just four days ago, ChatGPT introduced a new feature. Now it can listen, speak, and most importantly, see (Link):
It’s not yet available for all pro users; they are rolling it out gradually, but some fantastic examples have already emerged.
I’ve been using its listening capability for a while now. With the mobile version, there’s a voice-to-text option that works very well.
If you can’t wait any longer, you can use the plug-in available in GPT-4 called “Read that link.” Just input the link of the image you want to analyze in the prompt.
Honestly, it has improved significantly in recent months. It’s capable of interpreting an image and, from there, connects it with its vast knowledge of the world to derive insightful observations. For instance, analyzing this well-known chart to determine if it’s a good year to invest (spoiler: it is):
It’s not perfect, but it can indeed read text. And that opens up a plethora of possibilities. For example, I had it read the last page of George Orwell’s “1984”:
More examples: Send ChatGPT a SaaS screenshot and let it code for you (Link):
Extra Tip: If you come across a fantastic webpage, you can ask the AI to code it by simply providing the link.
And last but not least, I adore this example. ChatGPT will assist us in decrypting everyday problems.