{"id":3467,"date":"2026-09-11T10:56:48","date_gmt":"2026-09-11T17:56:48","guid":{"rendered":"https:\/\/www.armenshirvanian.com\/podcast\/?p=3467"},"modified":"2026-09-11T10:56:50","modified_gmt":"2026-09-11T17:56:50","slug":"469-ai-journals-research-and-better-systems","status":"publish","type":"post","link":"https:\/\/www.armenshirvanian.com\/podcast\/469-ai-journals-research-and-better-systems\/","title":{"rendered":"469: AI, Journals, Research, and Better Systems"},"content":{"rendered":"\n
\"\"<\/figure>\n\n\n\n

In episode 469 of The Armen Show, I talk through some of the ways I\u2019m using AI and other new tools in 2026, and what becomes possible when these technologies are applied to years of accumulated information, projects, and experience.<\/p>\n\n\n\n

One of the main examples is my journal archive. I\u2019ve been keeping electronic journals going back more than a decade, with my more recent material in Notion. Until recently, there was no practical way I was going to reread hundreds or thousands of entries and systematically extract patterns from them. Now AI can go through years of journals and help identify what actually worked, what repeatedly didn\u2019t work, which projects remained open-ended, which people and places kept appearing, and what conditions were associated with making real progress.<\/p>\n\n\n\n

That changes the value of keeping information. Saving something is useful, but saving it and then being able to analyze it years later is much more powerful.<\/p>\n\n\n\n

I also get into the growing role of AI agents and tools that can increasingly do the searching, sorting, organizing, and analysis themselves. Rather than learning elaborate ways of communicating with computers, the longer-term direction should be toward systems that understand ordinary communication and handle more of the intermediate work for us.<\/p>\n\n\n\n

Another example is research organization. I\u2019ve been moving papers into Zotero, where PDFs, titles, authors, metadata, and collections become much easier to navigate. Having a well-organized body of research changes what you can do with it later. The same principle applies to journals, computer files, books, podcast episodes, and other collections: once there is enough accumulated material and enough structure around it, the collection itself becomes a resource for future work.<\/p>\n\n\n\n

A broader theme running through the episode is compounding<\/strong>. A small improvement in how you organize information, use technology, communicate, or structure your work can save time or create opportunities later. Then the next improvement builds on that one. Over months and years, those differences can become substantial.<\/p>\n\n\n\n

Toward the end of the episode, I look at a few research ideas as well, including the illusion of moral decline<\/em>, which is the recurring tendency for people to believe society has become less moral even though generations have repeatedly reported the same decline, and the related tendency to remember the music, culture, and experiences of earlier periods of our own lives as unusually good.<\/p>\n\n\n\n

The larger question is what we can do now that we simply could not do before. The technology already available in 2026 creates an unusually large opportunity to revisit old information, organize existing work, discover patterns, and spend more time on the things that actually matter.<\/p>\n\n\n\n

Topics<\/h2>\n\n\n\n
    \n
  • Using AI to analyze years of personal journals<\/li>\n\n\n\n
  • Finding patterns in what works and what doesn\u2019t<\/li>\n\n\n\n
  • ChatGPT, Notion AI, and agentic tools<\/li>\n\n\n\n
  • Organizing research papers with Zotero<\/li>\n\n\n\n
  • The compounding effect of better systems<\/li>\n\n\n\n
  • Why trying new technology matters<\/li>\n\n\n\n
  • Building useful collections of work over time<\/li>\n\n\n\n
  • The illusion of moral decline<\/li>\n\n\n\n
  • Why people often think the past was better<\/li>\n<\/ul>\n\n\n\n
    \n