That ad headline you scrolled by or that sales email you received might have been written in part by an AI. Soon, just about everything we see and hear might have been partially penned by an AI. Although we may be far from AI independently creating complex creative works like a book or a movie, we’re just around the corner from AI becoming an instrument that virtually every creator uses.

Thought #1: We're already using AI

Inspiration: I don’t have hard numbers on this but I suspect people are getting more information not through human curated bundles of paper but rather algorithmically compiled feeds. Whenever we scroll through these feeds, our brains are loaded with ideas that percolate through our minds until we find the time to compose a thought.

Fact Finding: It doesn’t matter if you’re rich or poor, it seems like the number one research tool everyone reaches for is Google (or maybe Bing, you do you). The thing about search engines is that no one person really know how they work. These search engines are very much a black box of AI magic yet we trust them to find us information that we readily assemble into something worth publishing.

Proofreading: AI already powers many spell checking, plagiarism detectors, and grammar tools. Perhaps one of the most telling sign of how reliant people are on AI is that many Chinese natives have forgotten how to write Chinese characters. Since most writing is done through typing phonetic pinyin and choosing the characters to commit on screen, Chinese natives are losing the ability to write.

Thought#2: How generative AI can improve upon content production

Customized inspiration: This is possible because you can use GPT-3 within your text editor so it understands the exact context of the piece you’re creating. This way creators break through creative block by soliciting a few ideas from an AI whenever they’re stuck.

Increased output: AI can dramatically reduce the time it takes to write since AI is no longer an inaccessible feature of a search engine or social feed but now just an API call away. This means developers can embed AI directly into a document editor or a web app. Skipping the context switch of jumping into another tool will free up more time for creators.

Improved reader experience: While it’s easy to A/B test copy or a landing page, it’s hard to A/B test a blog post or other long-form content. But with AI tools today we can ask for a blog post and even the header image of that blog post to be adapted many times. Rather pick a canonical version of a story, it’s even possible to serve a personalized blog post to every visitor if we have context clues about what they like. Some might enjoy humor in their news, others may not.

Thought#3 and beyond

I've continued my writing about AI on All the AI, a newsletter dedicated to my generative AI obsession. Check it out to hear my takes on everything from how AI will affect Google and search to principles for building with AI.