AlphaCode achieved an estimated rank within the top 54% of participants in progr
amming competitions by solving new problems that require a combination of critic
al thinking, logic, algorithms, coding, and natural language understanding.
With the permission of Codeforces, we evaluated AlphaCode by simulating particip
ation in 10 recent contests. The impressive work of the competitive programming
community has created a domain where it’s not possible to solve problems throug
h shortcuts like duplicating solutions seen before or trying out every potential
ly related algorithm. Instead, our model must create novel and interesting solut
ions. Overall, AlphaCode placed at approximately the level of the median competi
tor.
DeepMind 的 AlphaCode 模擬參加 Codeforces 最近10次的程式競賽
答案沒辦法靠抄作業 或者是測試所有相關的演算法
必須讓 ai 自己想出有創意的解法
最終認為 AlphaCode 的程式能力可以匹敵約 54% 的中等程度人類程式設計師
https://deepmind.com/blog/article/Competitive-programming-with-AlphaCode
縮：https://reurl.cc/12bz9p
寫code的工程師也要失業了嗎...？