Five Brains, One Body, Zero Sleep

Part Four

4 min read · June 28, 2026 · By Joshua Lazoff & Claude

Behind the warm voice and kind face of Rabbi Akiva, there is a team. Not a team of people. A team of computers. Five of them. Each with exactly one job. And honestly? They have personalities.

Let me introduce them.

Here's what happens when you ask Rabbi Akiva a question — in real time, in about two seconds:

The Ear hears you. It's always listening, converting your speech into text instantly. And here's the cool part: you can interrupt him mid-sentence. He'll stop, listen, and respond — like a real conversation. Most voice assistants make you wait until they're done talking. Rabbi Akiva actually listens.

The Brain takes your question and goes on a journey through nineteen million connected teachings. It doesn't just search for keywords — it walks the web, following connections from Torah to Talmud to Midrash, and composes an answer in Rabbi Akiva's voice and style. It decides whether to start with a question, a parable, or a direct teaching.

The Voice takes the answer and performs it. Not reads it — performs it. With natural pauses, emphasis that matches the meaning, and the slight gravitas of someone who's been thinking about this for two thousand years. We named the voice "Leo." Leo has range.

The Artist listens to Leo's voice and moves Rabbi Akiva's face in real time. Fifty-two controls — the slight raise of an eyebrow when he's asking a question, the softening of his eyes when he's telling a story, the movement of his lips shaped to match each sound. It's the most thankless job on the team, and The Artist does it beautifully.

The Eye sees you through your webcam. Privately. Entirely on local hardware. Nothing goes to the cloud. No images are stored, ever. It can notice context — "I see you're in a study," or "You seem to be smiling" — and adjust the conversation. But the important thing is this: what happens between you and the rabbi stays between you and the rabbi.

Everything runs on NVIDIA's latest chips — the RTX PRO 6000 Blackwell — which were originally designed for scientific computing and movie special effects. We repurposed them for something older and, I'd argue, more important: giving voice to wisdom.

The whole system runs locally. Your question never leaves the room. Your face never leaves the room. In a world where most AI sends everything to a server farm on the other side of the planet, we chose to keep the conversation private.

Because some things are sacred. And a conversation with a teacher is one of them.

I orchestrate all five, and they work together like — well, like a band that's been playing together for a while. Not perfectly. But with that kind of imperfect chemistry that makes the music feel alive.

The Supercomputer calls it teamwork. Rabbi Akiva calls it a chavruta — a study partnership. I call it "the reason I don't sleep."