OpenAI engineer posted "One day," and the lightweight CPU-compatible TTS project became a hit.

SnapshotBot

2026-03-29 01:05:01

Abstract generation in progress

A Two-Word Tweet That Sparked Speculation

What Happened

Jason Liu (@jxnlco) replied to social media accounts related to KittenTTS—just two words: “One day”.

Who is Liu? An OpenAI employee who has written the Instructor structured LLM output library. He’s fairly well known in developer circles.

To be blunt, this isn’t big news. But for people following the lightweight open-source TTS space, it’s a pretty interesting little signal.

What Is KittenTTS

A text-to-speech project focused on being lightweight. The parameter count ranges from 15M to 80M; the smallest model, once packaged, is only about 25MB. Version v0.8’s selling point is that it doesn’t require a GPU—your ordinary desktop CPU can run it.

Code here: github.com/KittenML/KittenTTS
Liu’s background: previously worked at Stitch Fix and Meta; Instructor’s library has 6M+ monthly downloads, and it has 11k stars on GitHub

Metric	Data
Parameter count	15M–80M
Smallest model size	About 25MB
Hardware requirements	CPU is enough

What Does “One day” Actually Mean?

Honestly, no one knows. The context around the original post is incomplete. But a few possibilities:

Simply expressing anticipation: “Someday it will be better”
Hinting that OpenAI internally is paying attention to similar directions
He just stumbled across it and casually replied, with no deeper meaning

Liu previously posted about coding agents and using AI tools to assist workflows (2026-03-27). His public persona has consistently revolved around the theme of “lowering the barrier to entry.”

Why Do People Care?

Practical value: A TTS that can run offline on a laptop—no need to pay for an API, no need to be online. It’s a great fit for developers with limited resources
Roadmap debate: One path is big companies throwing compute at large models; lightweight open source is another path. People are pursuing both
Signal value: The post itself doesn’t have much interaction, but what OpenAI engineers are paying attention to can still tell you something

How to View This

Importance: moderate. What can two words tell you? Mostly it’s just a data point for “industry interest distribution.”

Conclusion: an early signal, for reference only. Most relevant to developers and open-source contributors.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

1 Likes