One mistake we made early was assuming that if someone clearly had the problem, they would be willing to change how they solved it.

Not necessarily.

A business can dislike its current process and still prefer it because it is familiar.

That changed how we think about product adoption.

I ve been thinking about what actually makes me uncomfortable about letting an AI agent take actions.

It may not be the size of the action.

It may be whether we can undo it.

I've been sitting with this question for a few weeks and I can't shake it.

If an AI agent can find the prospect, write the email, personalize it by vertical and persona, follow up three times, and log everything in the CRM - what exactly is a GTM team doing now? The easy answer everyone reaches for is "strategy." But AI can already A/B test messaging, read pipeline patterns, and shift targeting in real time. That answer gets thinner every month.

Over the past few days, you may have noticed a new account on Product Hunt called "Curious Kitty" commenting on some launch posts. It was created on 20th December 2025.

First observation: it behaves like a bot. It only asks questions (no opinions, no praise, no criticism) and those comments get visually highlighted, similar to how awarded comments stand out on Reddit. (screenshot attached)

Running agents in production is getting expensive fast especially when something loops, retries, or a user abuses the system. Curious what others are actually doing:

• Relying on provider-side billing alerts?

• Hard limits set on the OpenAI/Anthropic dashboard?

• Custom solution you built yourself?

• Nothing yet and just hoping for the best?

 I've been deep in this problem lately actually built something around it (launching tomorrow on PH). Would love to hear real approaches first though, especially from anyone running multi-tenant SaaS where you need per-user cost control.

Pulled my own kitty points breakdown again since it moved a lot since the last time I checked. Sitting at 388 KP total now. Forum Threads plus 272, Comments plus 101, Reviews plus 10, Maker plus 6.

Ten days ago Forum Threads were plus 87 and Comments were plus 64, so the gap did not just grow, it flipped which category actually drives the number. A handful of threads are now outscoring every comment I have ever left combined.

I'm putting together the gallery images for our launch and the recommended size is 1270x760, which is a laptop shape. Every guide says to bake the captions into the image. On a phone that same image is the width of my thumb.

So I'm designing for a screen I'm not sure anyone is using.

This matters to me as a marketer. The gallery is the first thing a potential user sees of the product, and if the type is unreadable on their format, the whole first impression is gone before they read a single word. Same image,

completely different experience.

Hey everyone,

I ve been talking to a lot of SaaS and marketing teams lately, and there s a massive recurring theme: everyone is using AI to write, but almost everyone complains that the output sounds generic and lacks personality.

When you give an LLM a prompt like "Write a blog about X," you're giving it a topic, but completely starving it of context. It misses the brand voice, the specific audience positioning, and the terminology that makes a company unique.

I'm curious how other founders and content leads are handling this friction right now. Are you manually pasting your brand guidelines into your prompts every single time? Have you built custom custom GPTs? How are you actually bridging the gap between generic AI text and real, brand-aligned content?

We all hear the same advice: "build in public," "launch on Product Hunt," "find your niche," "talk to customers."

But what's one piece of this advice you think is actually overrated, misleading, or just doesn't apply to your experience?

For me, it's "ignore your competitors." I think watching them closely is how you find gaps. Curious to hear yours.