I probably should not be surprised that AI could help with one of my personal headaches.
TL;DR: I wanted a better way to gather and process nuclear news, so I used AI tools to build a system that gets me to the good stuff. It turned out far better than I expected, and it was interesting enough that I decided to share it.
When persistent friction in a process keeps getting in the way of thinking, that is usually where technology can make the biggest difference. If we can use tech to take care of the little things that give us headaches, we give ourselves more space to make big things happen.
That mindset is always in the back of my head when I'm helping teams improve workflows, strengthen systems, or tackle persistent challenges. This weekend, I decided to apply it to my own life.
I love following nuclear news. There is something happening nearly every day that shapes the future of the industry. The challenge is not a lack of information. It is the volume. So many outlets echo, repost, or repackage the same stories that it becomes hard to see what is actually new. You can scroll through pages of headlines and still walk away with only a handful of truly distinct articles. Even worse, the real gems are often buried. At some point you slip into autopilot, skim past something important, and only realize it later.
You eventually narrow things down, but by then you are already nearing the end of your mental capacity. Either the day pulls you elsewhere, or you have spent so much energy filtering that you never fully process what you just read. There is often a deeper insight just out of reach. An idea that almost connects to something you saw earlier in the week but does not quite snap into place. You know that if you could come back with fresh eyes, something might click. But by the time you do, a new batch of news has arrived and the cycle starts again. I feel the headache coming on just thinking about it.
I tried the usual solutions. Keyword alerts. RSS aggregators. Limiting myself to one or two trusted outlets and relying on their curation. Every option had a downside. Too much noise. Too many gaps. Or too much process just to get to the information I actually cared about.
What I really needed was an assistant. Something that could look at everything, pull out the stories I would find interesting, dig deeper on those articles, summarize them, highlight key indicators, and help surface connections across the week. And still let me get to the original sources when I wanted to go deeper.
So I decided to try building it.
I am not a coder. I have some familiarity with programming concepts from classes taken long enough ago that we would need to count in decades, not years. But I had been hearing more and more about how AI tools were helping people build practical systems. I wanted to see if that capability could help solve my nuclear news problem.
I opened ChatGPT and started talking. I described the issue, asked questions, and began assembling something I never would have guessed I could create. It was my thoughts translated to code instantly. If it wrote something I didn't understand, I could just ask for a line-by-line explanation. Sometimes I would get into the weeds on libraries and programming structure. And occasionally I would trust the process and let the results speak for themselves.
Piece by piece, a system came together.
In the end, I had exactly what I was hoping for. A pipeline that gathers sources, skims and grades articles, performs deeper dives on the most interesting items, and produces a short weekly brief that connects themes and trends across the nuclear landscape.
It removed a layer of friction that had been getting in the way of how I wanted to think about nuclear news.
And honestly, I thought it was worth sharing.