Is AI-generated outreach making cold email worthless or just making bad cold email easier to spot?

I've been on both sides of this lately, and I'm honestly not sure what to think anymore.

On the sending side, I've used AI to build outreach sequences. The copy was tighter, the personalization tokens were in place, and open rates looked fine. Replies, though, were thin. When I looked at who we were actually sending to, I realized the targeting logic mirrored the same blunt pattern-matching the AI used to write the message: job title, industry, company size, done. Technically personalized. Strategically random.

On the receiving side, my inbox is full of emails that know my name and reference my LinkedIn bio, yet still miss my actual situation. The effort is visible. The relevance isn't.

What I keep coming back to is this: the emails that still land aren't the ones with slightly better copy. They're the ones where someone clearly thought about fit before hitting send. Not just "does this person match the ICP?" but "does this specific person have a reason to care right now?" That kind of reasoning isn't something you can automate with the same tool you used to write the subject line.

So maybe AI outreach isn't killing cold email. Maybe it's just making the underlying laziness more visible and more scalable.

Has AI outreach changed how you think about who you reach out to, or only how you write the message?