Demystifying Generative AI in 10 Short Essays
Useful succinct insights to read with your morning coffee
I’ve wanted to write a themed series since I started The Algorithmic Bridge. Today I bring you an idea that combines that wish with a necessity that’s emerging out of the deafening noise and the unsettled dust: Demystifying (the present of) generative AI.
In a matter of weeks, ChatGPT popularized AI to heights never before reached during 70 years of research. Yet popularity doesn’t equal intuitiveness or simplicity.
Besides ChatGPT, people can use Bing, Bard, Claude, and a myriad of customized chatbot variants, not to mention AI art models like Midjourney and Stable Diffusion and other systems that don’t fit into those categories. Yet tool usability and manipulation precede understanding.
We’re living in a constant tidal wave of overwhelming information. Yet quantity doesn’t mean clarity.
Confusing, conflicting, and controversial takes abound. How can anyone know what to believe? Generative AI is too new and moves too fast. The leaders we listen to mix science with marketing, facts with predictions, and fears with hopes. With this “Demystifying Generative AI” series I intend to draw a clearer—albeit certainly not perfect—picture of critical topics that remain under debate and without consensus.
The articles will be short and light (hopefully around 3-4 minutes of reading time). Each will cover one specific question, without generalities. I’ll avoid hypotheticals and forecasts and focus on facts backed by evidence. I won’t approach them merely as critiques or appraisals but as realistic—and opinionated—takes.
I want this resource to be what I imagine I’d be looking for if I was now getting started with generative AI as an outsider. I hope you learn something along the way. I’m surely learning while crafting these articles. Despite the time I’ve been studying, reading, and writing about AI, I’m still trying to get my head around everything that’s happening.
I want to ask you one favor: Let me know in the comments if you like this format. (If you prefer long in-depth essays like the ones I usually share, don’t worry; I won’t stop writing those anyway.) Once I finish the series, I’ll link all articles in one place for easy access.
That said, I leave you with the first issue!
Generative AI Is Not a Magic Wand
There’s a gap between what we imagined we could do with generative AI and what we can actually do. A divergence between expectations and reality. François Chollet says that people love to imagine what AI can create but “not so much the actual reality of using it.”
This wasn’t at all clear a couple of months ago. Only firsthand experience with state-of-the-art tools contrasted with our preconceptions provides such a powerful insight.