Their crawlers are more advanced now. They already handle SPAs with # URLs and content that’s only fetched once the JS is executed, and likely use some OCR and more powered by AI magic.