Introducing CodeAlchemy, a synthetic data pipeline that has already produced nearly 1 trillion tokens of open-source code

We spoke with IBM’s David Cox about the lab's evolution and what it takes for an industry-academia collaboration to succeed.