teknopurge
Joined 9 years ago
Joined 9 years ago
If the value is in better training data then everyone will train their own models. We should have a goal where we have domain-specific models that make us more efficient, but also constantly train a context-aware model based on individuals. Model composition is the future.
> These accounts were not used to accept user deposits; they were used to facilitate user purchases” of crypto, said spokesperson Brad Jaffe. “There was no commingling at any time because these are 100% corporate funds.” When users sent money to the account, he said, they were not depositing funds but buying the exchange’s bespoke dollar-linked crypto-token, BUSD. This process was “exactly the same thing as buying a product from Amazon,”
The article has a lot of details and infographics - but ends with no conclusion, just guesswork. I appreciate the work and time that went into an article like this - it would be nice if it was more factual and less "blind-sourced" extrapolated hypothesis.