My new Nothing Ear earphones had been working great, but after a few weeks I noticed that sometimes they sounded like they were underwater. The audio would be extremely quiet and muffled. No matter how I adjusted any of the EQ settings, noise cancellation, in-ear detection, codec choice, or anything else in the Nothing app, the problem persisted. A full reset of the earphones didn’t help. I reached out to Nothing support and heard crickets to this day and decided I’d give it a week or two more before returning them.

I got an Akko MOD007B v3 a while ago and this weekend swapped out the PCB for the flex-cut version.

I’ve been exploring ElevenLabs TTS for synthesized audiobook narration. ElevenLabs have recently announced a community initiative encouraging discussion about the project (the incentive being TTS character allocation). Since this post has been simmering in my head for a while anyway, I figure now’s a good a time as any.

I visited Greece last year, spending a week in Athens and then another in Thessaloniki to attend a wedding.

I took so many pictures and they’re just languishing in my phone, so I decided put them into a small travel post.

There are many ways to use your own books in AI workflows. In this post I’m going through the most basic workflow that does not involve coding your own app, creating vector stores, or generating API keys. You can do this by just using ChatGPT (with a Plus subscription). Many (most?) writers are not programmers, so there is no coding required here.