Bytedance argues most AI training phases are unnecessary and backs it up with a competitive open model

Media reports The Decoder Source language: English

Bytedance and partner US universities claim that most AI training steps are redundant, suggesting only three phases are needed. Their new paper presents a competitive open-source model to validate this simplification, aiming to cut costs and improve reproducibility.

Full text

Jonathan Kemper View the LinkedIn Profile of Jonathan Kemper Nov 21, 2026

Plus AI research

Bytedance argues most AI training phases are unnecessary and backs it up with a competitive open model

Nano Banana Pro prompted by THE DECODER

AI agents increasingly rely on complex training pipelines. A team from TikTok parent Bytedance and several US universities aims to show that three steps are enough.

AI models are expected to research the web on their own, run code, and solve multi-step tasks. According to a new paper, the training methods behind these capabilities are growing more and more nested, with many developers stacking numerous training phases one after another. The paper argues it's become almost impossible to tell which step actually helps, and reproducing the work is getting expensive.

THE DECODER+

Sources

← Back to Daily AI Report