I recently put my old frontend hat on and opted for Framer to build a new website, seeing if I could sketch it out in GPT + Claude then had it give me instructions to translate to a modern cloud CMS.

While I find Framer, Webflow, etc, significantly more aligned to my brain than say Squarespace or Wix, overall, they are incredibly frustrating and full of tedious little proprietary quirks that drive me goddam insane. You need to learn the interface but still also need an understanding of how breakpoints and CSS positioning work among other things.

The preview website artifact that Claude Design spat out was pretty dang good and was written in simple vanilla semantic HTML/CSS with no dependency overhead. Its a little rough around the edges and clumsy on details but an appropriate starting point for both beginners and enterprise.

A company like webflow has to be looking at this and sweating bullets. Wish the initial reaction to AI-driven threats these days wasn't to gut the company in favor of agents. Yes Claude Code + Claude Design are awesome but none of these products are perfect and they all still have a place to coexist within the market

https://www.gurufocus.com/news/8881760/wix-plans-major-layof...

Never used either, but these are similar “no code” platforms, right?

I work in exactly this field and I make relevant revenue. I absolutely do not think this is in any way true nor proven by actual needs. In fact this is not even happening as described. Companies like webflow are following blindly an empty promise with the hope at some point it becomes self-fulfilling, disregarding their people and sacrificing any morals on the altar of growth for growth’s sake and investors pressure. All this to say, more concisely: fuck Webflow as well.

Leaving implies agency. You've fired these people. Using a euphemism that implies they had a choice in this matter is disingenuous and not "owning this decision"

What a disgusting way to phrase that.

But I am a webflow customer looking at ripping it out of our stack and just having it maintained by our team with alternative tools, powered by Claude/Codex.

I planned to follow some flavor of this -https://x.com/anitakirkovska/status/2053941736049967285?s=20

We currently pay almost $5k/year for hosting/licenses and its just not feeling worth it.

(disclaimer - still believe Webflow is 10x Wordpress!)

Interesting that it's not the development team that gets to start with using AI agents (unless they are already?), but instead the marketing team. I guess long-term quality matters less there, you just want something OK fast, then LLMs actually makes a lot of sense, it is basically what we've ended up calling "slop" anyways, and marketing seems like a natural fit.

Interesting contrast compared to "I think Anthropic and OpenAI have found product-market fit" from the other HN submission right now, as that one claims mostly software engineers are using agents, meanwhile Webflow starts with the marketing team, I'm guessing they see it as obvious somehow.

Translation: Webflow has achieved AGI internally.

We've raised a lot of money and VC's are pushing for exit. It's now or never. Rebrand, slap on agentic AI and pray to google gods for a quick buyout.