Every few months, we get a new model that's faster, cheaper, and better than the last one.
But I've come to realize that the AI itself is not the most important bottleneck anymore.
It's the ecosystem of things around it that gets more difficult designing a workflow that feels intuitive, building trust, reducing latency and making sure that the product feels like something the person will love to use every day.
I spent nearly four years on one idea before I finally pivoted to what I'm building now (@Basedash: AI data analyst). What mostly bothers me isn't that I was wrong, it's that I still can't tell you the exact signal that should've made me quit two years earlier.
At the time everything felt like progress. We had users, we had encouraging conversations, we had the occasional good week that convinced me the next one would be better. None of it was a clear "stop."
About a week ago, there was a case in my country where a woman took a Bolt taxi, which, instead of taking her to the designated place, started driving her to the Austrian border. In the end, the woman jumped out of the car to save herself.
[So the police were not very active, and Bolt just used only AI-generated answers, the company then announced it was a mistake in the GPS... but to be honest, I would say it was a kidnapping attempt].
The AI tool graveyard in my dock is embarrassing. I've downloaded and abandoned more "game-changer" tools in the last year than in the previous five combined.
But a few have genuinely stuck. The ones I still use daily without thinking about it:
Claude for anything that needs actual reasoning or long context
- Cursor for coding - it's changed how I write code, not just sped it up
- Perplexity for research that I'd previously just Google-and-pray
It is a question of choosing between two evils for us now. Neither option is completely free of flaws.
Human: Recruiters with "gut feelings" who harbor unconscious bias. they reject excellent candidates who just didn't go to the "right" school or didn't just "click." Inconsistent, unfair, and un-auditable.
AI: Algorithms whose training datasets are themselves replete with historical biases. They increase the scale of discrimination at light speed, becoming so-called black boxes that end up rejecting qualified candidates for reasons that humans cannot even fathom.
We are truly deciding to exchange messy, subjective human prejudice for cold, ruthlessly efficient algorithmic prejudice. Is that really an upgrade?
I d like to discuss how launch-day events and sponsored leaderboards can create fair discovery opportunities for smaller makers.
When hundreds of products launch on the same day, teams with established audiences can generate early votes and comments. For newer projects without that audience, being placed low in the leaderboard can result in very little organic visibility. In my case, my launch received only three click-throughs to my website from Product Hunt over most of the day.
OpenAI Day stated that five standout launches would receive $10k in API credits. A few days after the event, I still have not seen a public announcement of the winners or an explanation of how they will be selected. Is there a published update or timeline that I may have missed?
Would mechanisms such as rotating featured projects, category-based discovery, or a separate editorial/judged track make future events more useful for smaller teams?
