The video in the article explains it pretty well. In essence he over-valued IBM and then ignored their decline for years, even while the company's core business continued to dry up. He finally just exited, taking huge losses in the process.

The problem with value investing is that it works great for traditional companies with normal products, but many of the fundamental analysis models fall apart with "tech" companies who's largest asset may be the brand itself.

For example Uber, they own few vehicles, have few employees, and no exclusivity. How do you run a fundamentals analysis on that? A lot of traditional models would tell you that company is worth near nothing.