BREAKING NEWS from the Financial Times:
This seems plausible. DeepSeek may well have broken OpenAI’s Terms of Service and distilled1 their intellectual property without permission.
OpenAI may well have done analogous things to YouTube, New York Times, and countless artists and writers.
Karma is a bitch.
§
Bonus track. Here’s me discussing some of the many threats OpenAI, yesterday on CNBC’s Squawk on the Street:
§
In short, to recap: a company that made its name regurgitating and recombining sliced-up bits of intellectual property in statistically probable ways without due compensation is now threatened by … another company apparently doing the same, at lower cost.
Gary Marcus can’t keep writing new bio lines every time some new episode in the OpenAI drama makes him laugh.
To a first approximation, distillation does for models what scraping and training have been doing for the web, viz grabbing a bunch of data and compressing the data with neural networks.
OMG 😱. The Chinese would violate terms of service and steal data from someone who stole data?????
Actually laughable his outsize hubris he can't recognise the irony of claiming copyright infringement