@@ -108,16 +108,16 @@ interviews][Hard Fork: AI Action Plans], summarized what I was _supposed_ to do.
108108> before, solve the problem, and be able to regurgitate the memorized solution,
109109> without acting like you have seen the problem before.
110110
111- In 2025's competitive tech hiring market, where laid-off tech workers outnumber
112- available positions, some candidates spend months grinding leetcode.com to reach
113- the top 1%, so they may be lucky enough to have already seen the 1 interview
114- puzzle close to the 1,000 they could be given. Other candidates use cheating
115- software like Lee's. Meanwhile, experienced developers who could do the actual
116- job well but haven't memorized these puzzles struggle to complete them under
117- pressure. The unqualified candidate aces the test. The qualified candidate fails
118- to finish in time, though they can explain what they _ would_ do next, given more
119- time. For your overloaded hiring pipeline, which candidate would you advance?
120- The test ace or the struggling explainer?
111+ Some candidates spend months grinding leetcode.com to reach the top 1%, so they
112+ may be lucky enough to have already seen the 1 interview puzzle close to the
113+ 1,000 they could be given. Other candidates use cheating software like Lee's.
114+ Meanwhile, experienced developers who could do the actual job well but haven't
115+ memorized the puzzles struggle to complete them under pressure.
116+
117+ The unqualified candidate aces the test. The qualified candidate fails to finish
118+ in time, though they can explain what they _ would_ do next, given more time. For
119+ your overloaded hiring pipeline, which candidate would you advance? The test ace
120+ or the struggling explainer?
121121
122122Lee answered whether he had ethical concerns about his product.
123123
@@ -129,11 +129,9 @@ Lee answered whether he had ethical concerns about his product.
129129Ethics aside, AI-assisted coding tools are great at solving LeetCode problems.
130130They have memorized 1,000s of such puzzles. If something is easily solvable by
131131computers, maybe that's another hint it's not a great test of a software
132- developer?
133-
134- While I haven't subscribed to Lee's product, I know I like using AI-assisted
135- coding in my normal job. It would be realistic to have normal tools available in
136- an assessment.
132+ developer? While I haven't subscribed to Lee's product, I know I like using
133+ AI-assisted coding in my normal job. It would be realistic to have normal tools
134+ available in an assessment.
137135
138136While the debate of allowed tools in an assessment rages on, to skip the debate
139137next time I'm on the job hunt, I hope to cross-reference my applications with
0 commit comments