Sounds simple enough. You add it to the backlog, estimate the work, and move on.
But what if they don't actually need an export button?
Maybe they're copying data into another tool every Friday. Maybe their manager needs a report. Or maybe they just don't trust that the data will still be there next week.
These days, almost every product that launches comes with some form of AI. It's become the default AI for this, AI for that. And honestly, most of them don t really need it. The result? Everything starts to feel the same. The only real selling point becomes we use AI.
That s exactly why I started building @HumanEye because not every problem should be solved by AI. Some things, like resume reviews and career guidance, still deserve the human touch. Real feedback, from real people.
Would love to hear your thoughts:
Are we overusing AI just for the sake of hype?
Have you come across products that felt forced because of their AI features?
What are some areas where human input still matters most?
We route across a handful of models and on a spreadsheet the savings look obvious. The small model gets it right most of the time. The problem isn't the hit rate, it's the shape of the misses.
When a frontier model is unsure you can usually see it. The output hedges, goes vague, asks a question. When a small model is unsure it produces something clean, confident and wrong, and that sails straight past any eval checking format instead of truth. Nobody catches it until a user does.
Our tiny team got completely flooded with support tickets during our last major update. Critical bug reports ended up buried under simple stuff like password resets and basic account questions.
Setting up basic auto replies for common issues and organizing tickets by tags made a huge difference. It cut our response times way down and saved us from hitting total burnout.
I just pulled our analytics for October, and I feel sick looking at it. We finally cracked the code on traffic our SEO efforts and a few viral TikToks brought in 50,000 unique visitors last month. I was thrilled. I thought we were about to hit our first $10k MRR milestone.
Then I looked at the actual conversions. Out of those 50,000 people, only 40 created an account. That s a conversion rate of 0.08%.
I ve spent the last week analyzing session recordings, and it's painful to watch. People are landing on the page, scrolling around for about 15 seconds, and then leaving.
I showed the page to a friend, and he pointed out the obvious: our headline is way too vague. We wrote it for ourselves (we know what the product does), but a cold visitor has no idea what problem we are actually solving. Our call-to-action (CTA) is also buried at the bottom. We were so focused on getting traffic that we completely neglected the message for the people arriving.
It s so easy to get addicted to looking at positive analytics and charts that go up and to the right. I remember celebrating a bump in feature clicks, only to realize later through a quick user call that people were just clicking around out of pure confusion because the layout didn't make sense. The metrics counted it as high engagement, but reality was completely different. Have you ever trusted a green dashboard metric only to find out later that the actual user experience was broken?
