Pulling the future into the present 💫

GenAI redefines what "average" means

Wix released a preview of their upcoming AI witchcraft this week, and I’ve got to say, it’s pretty sleek. Type in a prompt and 💥 boom 💥 you have a website. Good riddance to the days of spending $49 on a website template only to bang your head against it until it breaks so you switch to Webflow which makes things even more confusing so then you make a pitstop on Twitter to complain and then hear a bunch of recommendations to use Framer, but you’re already too deep in this so you keep going because being a (no code) hacker is part of your identity and then eventually you pay someone $10 on Upwork at 2:34AM to fix everything in five minutes. (deeeeeep breath).

Anywho, to be honest, we’ve all been waiting for this, right? No-code website builders are incredibly powerful, but giving users the ability to infinitely customize means dense UXs that require joining a university to learn and navigate the product. This is a perfect use case for natural language prompts: instead of endless pointing/clicking and searching for certain components, I can just type out a request and be amazed.

To manage the difficult onboarding before GenAI, these platforms launched marketplaces of templates, specifically targeting the “ballers on a budget” who can’t (and won’t) drop $10k but still want a professional-looking website. While the templates were far from perfect, they got the user 60(ish)% of the way to the finish line. Moreover, they created an ecosystem of experts on their platforms that could develop endless variations of content ideal for the endless needs of their users.

The problem with templates is, of course, they’re templates. They’re average. Your story, brand, and platform now look legitimate, but it probably definitely doesn’t stand out. Your users likely aren’t having a religious experience on your website. However, the good news is that because of the template, you no longer have a bad website. You’ve been promoted to an “average” website! And that’s progress.

If Wix’s “coming soon” AI features end up anywhere close to the experience they demoed, then this is a huge success for the entire GenAI community and, more specifically, those bullish about the user experience of “natural language commands” to power an interface. Similar to what we’ve seen with Adobe Firefly, this reinforces the potential to improve all of the other complex no-code design experiences in our day-to-day: GSheets (think building a visualization), GSlides (hi, Tome), Figma (hi, Diagram), and so many more.

However, let’s be real: these products won’t be “fully autonomous” for at least another few years. These new AI features (1) enhance the template ecosystem by not needing to rely on experts to create everything and (2) make trialing the product easier. The underlying function of the tool has not changed. We’ll still likely need to get our hands dirty with endless point-and-clicks and custom code to solve edge cases, but it’s still progress.

Making complicated design tools easier to use means the digital world will look better / sleeker / more beautiful. I’m all for that, and it’s cool to see the future being pulled into the present by these incredible companies. In the long run, our entire definition of quality will change: as other platforms launch their own AI features, the quality we expect out of any given website will simply be higher.

Except Craigslist of course. I hope they never change.

Happy weekend,

Raman at Rhetoric

---

📚 What’s made me a better storyteller this week

Dan Frommer just dropped his latest New Consumer Trends report (with a focus on GenAI, Ozempic, and finances). It takes about 15 minutes to flip through this and is a great exercise in unbiasing yourself from your own assumptions about the world.

I get a lot of questions about the best way to “catch up on the basics” of LLMs. I thought Cohere’s LLM University did a great job of explaining (visually) the most important concepts.

Turns out big top-line revenue numbers are taking longer to translate into real-world money-in-the-bank. Here’s a good read by David Heinemeier Hansson: “it also means that there's probably a lot more pain in the pipeline for unprofitable software startups, almost all of which are now SaaS, which raised money in the late go-go days.”