It’s been a while since I’ve written about a sea change in technology. I think one of the last topics was the rise of Big-Data where collection, analysis and compute resources were all coming together to fuel a revolution in predictive analytics which we all take for granted today. For instance, we get upset when Google maps doesn’t re-route us even when the backup just happened. Or how we now run genomic sequences on our dogs.
We’ve been spoiled.
I’m now seeing the same momentum, hype, and emerging reality
around Gen-AI both for large language models
and Image-AI. We stand on the cusp of
Gen-AI changing the way we learn, create, and make decisions.
The concepts of Gen-AI are not new mind you, Microsoft
rolled out AutoCorrect in MS-Word in 1993.
Apple released Autocorrect with the debut of the iPhone and IOS 1.0.
What’s different is that today’s perfect storm of parallel
processing GPU’s, (Related
article Microsoft’s own ARM based GPU, look out Nvidia) combined with the neural
networking algorithms that can correlate large associations of known language
and image results and the open source large language models e.g. Open-AI with
its 175B Parameters, all combine to fundamentally change the way we generate
content and “original” thought.
Open-AI’s ChatGPT, one of the most public and widely used LLM’s stands to either be a breakthrough in the way humans create content or a shortcut
to homogenize and normalize the world around marketing doublespeak, depending
on your point-of-view.
Personally, I like the way Google Pixel Image-AI helps
everybody take a great picture.
Big-Data is all about using large data sources to form
better predictive analytics and facilitate decision making to augment human
action that drives better outcomes.
Gen-AI is a giant leap in the same direction, using those
same Big-Data concepts but to change the way ideas are formed and content is created.
The dilemma is that this is the essence of creativity and
what makes us human.
