A retrospective · ~5 min read
For roughly fifteen years, "public data" and "your data" felt like the same thing on Twitter. In February 2023 they stopped being. The lesson wasn't about pricing — it was about what "public" is actually worth.
February 9, 2023, Twitter announced free API access would end. March 30, the "basic" tier launched at $100/month with rate limits so tight most tools couldn't run inside them. April, third-party clients Twitterrific and Tweetbot both shut down — their business models presumed the API would keep existing at rates a small team could afford. Twelve years of ecosystem work, gone in six weeks.
The archival tools were quieter but hit harder. Nitter instances, tweetdeleter services, the community backups people had built to hedge account suspensions — most of them broke and never recovered. If your history hadn't been personally exported by early 2023, you were now dependent on Twitter's own "Your data" download button being available on the day you needed it.
The core misconception the 2023 changes exposed: that "publicly visible" implies "publicly retrievable." It doesn't. A tweet you posted in 2015 is visible to whoever loads the page today — but retrievable by a tool that wants to help you keep it? Only if the platform lets that tool exist. Access is a policy, not a property of the data.
This isn't a Twitter-specific problem. It's a property of anything you've entrusted to a platform for archive-by-visibility. The correct backup is not "it's still on the site." The correct backup is a file on your disk.
Konshus is one way to solve this — a persistent memory vault and portable persona that follows you across ChatGPT, Claude, Gemini, and whatever ships next.
Meet KonshusIronically, the personal "Download your data" flow — Settings → Your account → Download an archive of your data — still works and produces a complete ZIP: every tweet, every DM, every media file, every follow-and-follower list. It's a user-facing feature and it's cheap to run, so it survived the ecosystem purge.
Getting the archive is the easy part. What to do with it is the harder question. A ZIP on your Dropbox is a backup; a ZIP that feeds a memory tool that can quote your own writing back to you is more.
Konshus's X importer takes the ZIP, filters retweets and stranger-replies out, and imports each of your original tweets as an artifact. Default subject is "about_me" — your voice, in your words, across whatever era you were writing. Waitlist at /vault/import/x-archive.
Ten years of your original tweets are usually a better voice sample than most journal entries, because they were written to be read — you found your public voice out loud. Every model that quotes your Konshus back to you gets to reference that voice.
X won't stay public forever. Request the archive, drop the ZIP, keep the voice you built.
Meet Konshus