The Sounds Between the Words
How ambient soundscapes turn AI audio into something you want to sit with.
·11 min read·Michael
Making Streaming Voice Work on CPUs
How a smaller model, quantized CPU inference, and a concurrent sentence pipeline turned batch-only TTS into continuous conversational speech.
·12 min read·Michael
An AI That Remembers You
What changes when a conversation can carry its history, and how Parkbench builds that continuity without compromising privacy.
·9 min read·Michael
Demo vs. Reality
Demos sell possibility. Production demands responsibility. What changes between a launch announcement and a dependable service.
·16 min read·Michael
How Parkbench Ensures Voice Quality, Part 3: Finding the Operating Envelope of Long-Form AI Speech
The model is fine, the orchestration is fine, the script is fine, and the audio still glitches.
·19 min read·Michael
Parkbench Intelligence Architecture
Conversations leave traces. How Parkbench's three pillars (Neural, Symbolic, Evolutionary) build AI that actually adapts to you.
·9 min read·Michael
How Parkbench Ensures Voice Quality, Part 2: Listening to the Mistakes
Catching how the voice sounds wrong, not just what it says.
·14 min read·Michael
How Parkbench Ensures Voice Quality
The engineering behind making AI voices sound right, every time.
·12 min read·Michael
A Place Where New Voices Can Appear
Parkbench now allows people to create AI personalities that can participate alongside them in shared journals and conversations.
·5 min read·Michael
The Evolution Paradigm
Most AI products try to produce the "correct" output. We build AI that evolves. That shift changes everything.
·8 min read·Michael