They train LLM only with data up to 1930 and it still solves Python problems (theregister.com) 4 points by Multipassionate 4mo ago ↗ HN
1 comment
[ 3.7 ms ] story [ 16.9 ms ] thread