Stop Optimizing Your Cold Emails. AI Has Already Ruined Them.

You open your inbox, and there it is. Another perfectly structured, flawlessly punctuated, aggressively polite email from someone you’ve never met. It has the hooks, the value propositions, the seamless transitions. It is a masterclass in copywriting. And you delete it without a second thought.

We used to love getting cold emails. They felt like a serendipitous connection, a real human reaching out across the digital void. Now? They feel like a threat. Because we know exactly what’s on the other end: not a person, but a prompt.

We didn’t lose cold email to bad writing. We lost it to perfect writing.

The paradox of efficiency is finally catching up with us. Automation tools and large language models have made cold emailing infinitely scalable. Anyone can generate 10,000 personalized-sounding emails in an afternoon. But the very tools that help senders have trained receivers to distrust any message that feels too polished. You read the same generic fluff, the same predictable sentence structures, and your brain instantly flags it as noise.

I read a dozen inbound cold emails every day. I used to engage with them. I don’t anymore. Because every other single email uses “The Sentence.” You know the one. That overly enthusiastic, slightly unnatural phrase that no human being has ever actually spoken out loud. Whatever model these people are using, it always puts in The Sentence. And the moment I see it, I’m out.

The implicit social contract of email has been broken. We used to assume that if someone took the time to write to you, they actually wanted to talk. Now, we assume they just ran a script. Trust has become the scarcest resource on the internet.

Most people try to fix this by optimizing their cold email copy even more. They tweak the subject lines. They A/B test the call to action. They beg the AI to sound “more human.” But this is a losing game. You cannot out-human a machine by smoothing out the edges.

When everything is optimized, the only way to stand out is to be unapologetically flawed.

The real battle isn’t in the copy; it’s in proving you’re a human in a way an algorithm can’t fake. AI is designed to seek perfection, to correct typos, to structure thoughts logically. To beat it, you have to do the opposite. You have to introduce friction.

Leave a deliberate imperfection. Send an email from a medium no AI would choose. Write a paragraph that is too long, too specific, and too weird for a bot to generate. Reference a hyper-niche detail that proves you actually looked at their work, not just scraped their website. I got a cold email recently and was actually happy to engage, but I warned the sender upfront that LLMs change the game. I asked for more info about their background to authenticate them as a human. They replied with another unnecessarily long email that was light on details and reeked of AI voice. I ghosted them.

If you want to land a conversation today, you have to be willing to be messy. You have to be willing to fail. Because the worst thing anybody can say is no—and in the age of AI, they’re already saying no to anything that feels synthetic.

AI didn’t just break the social contract of email; it made deliberate imperfection the new authenticity.

Stop trying to scale your outreach. Stop trying to sound like a Fortune 500 CEO. Be a weird, specific, flawed human being. That’s the only thing left that’s worth buying a lotto ticket on.