Google has 2.8k public repositories on just their main github account (github.com/google).
Even if they're not training on their closed source internal codebases (which they most certainly are, but fair point that they're probably not releasing the models trained on that data), it definitely seems like they have a lot of Go code in the training data.
Just means it's better at one specific task than the others, which has always been the case. For each of Sonnet, GPT and Gemini I can readily name a task they are individually the best at. At the same time the consensus that Sonnet 3.5 is overall the currently strongest model remains correct, and that's what most people care about. Additionally most people do tasks that all of the models perform similarly at, or they can't be bothered to optimize every task by using the best model for that one task. Which makes sense since not a single cloud provider has all three of them. Now this one will likely be AWS-exclusive too.
Benchmarks are way too easy to game. There's no shortage of models that "beat GPT-4" according to some benchmark or another, that are obviously nowhere even close when you try them on novel tasks.
Can't read the article due to paywall, but potentially ambergris? It's a form of whale excrement that washes up on the shore and can sell for ~$10k USD/pound
May be resolved, or just spotty? I was able to load the page itself but tweets and any trending information were all out, seeing a lot of similar reports
