Last couple of days, a website called Podcast.ai went viral amongst my friends. 

The site, powered by text to voice generation tool play.ht, generated an ‘interview’ between Joe Rogan and Steve Jobs, entirely by AI. 

It is not perfect – but I have to say, it is pretty impressive. If you have heard Steve Jobs speak (on YouTube perhaps) a few times, you might not be able to tell the difference. 

Also, over the National Day holidays (1-7 October), I have been playing with Stable Diffusion, a tool that turns text descriptions into images. The results are actually quite impressive as well. 

Stable Diffusion is not the only text to image generator out there, you also have DALL.E 2, Midjourney etc. 

Somehow, I feel, that we are at the cusp of artificial intelligence – at the current pace of evolution, good commercial grade applications might emerge very quickly. 

What killer applications will there be? I do not know – probably nobody knows. But that’s the beauty of it – with an enabling technology and so many different people trying different things, interesting applications that you never thought of will emerge. 

Some friends of mine are already building some of the algorithms into their tools – allowing game designers to narrate a story for AI to render automatically for example. 

And Google is also enabling text to video. 

Exciting times!