The challenge is all legal. Every single AI shops is training on pirated data corpuses and grey area scraped data so obviously not a single company is willing to reveal their data.
The other countries don't have cheap capital, cheap infra, or scale. The talent is not the bottleneck. Look at how many SV companies are helmed by Indian executives. If the Europeans cared an iota about AI progress, Huggingface wouldn't have been headquartered in NYC.
This behavior is actually specific to ChatGPT because they lost a music copyright lawsuit in Germany. They would refuse to output music lyrics too but they would happily do analysis on lyrics if you supply them. I suspect there might be a guardrail model involved here.
I agree, it's the old academia play. When they can't innovate on the model work, they make a lot of noise about AI safety, gatekeeping, humanity extinction etc.
>Alongside the original snapshots, we’ve included a few AI-powered modernization attempts that demonstrate what’s possible—getting this 1990s-era C++ and MFC code building with current Visual Studio tools, connecting to modern IRC servers, and running legibly on today’s high-resolution Windows machines.
Given that MSFT is all in on Rust and WinUI now, maybe they can try doing a full port similar to Bun using Copilot. Anthropic has been milking their Bun port attempt for as much as they can.
I disagree with the approach. It's a good approach for limited domain problems, but not for general purpose. Take something like this where you will need to be able to refer to wikis and research and ask questions on Reddit and Discord to optimize playthrough, none of the goal conditioning will be useful.
Microsoft can probably scale Minecraft up to support infinitely large sized worlds if they want to. Minecraft lazy loads the world in chunks so you there's no real limitation (aside from machine digit sizes)