Nate Soares has written a post about the common mistakes everyone makes when coming into AI alignment.

Judge as of 2023-04-06. True: a post written by Nate Soares and published on or before 2023-04-06 (e.g., on LessWrong, the Alignment Forum, the MIRI blog or his own blog) addresses the common mistakes or failure patterns of people new to AI alignment. Mostly true: such a post exists, but its focus is somewhat broader or narrower than 'mistakes newcomers make' (e.g., how most alignment plans miss the hard parts), while still clearly covering that theme. Mixed: Soares wrote on related topics, but no single post substantially matches the description. Mostly false: the matching post was published after 2023-04-06, or was written mainly by someone else with Soares only a minor contributor. False: Soares's publication record up to that date shows no post on this topic.

Set before the research began.

1. E1 On how various plans miss the hard bits of the alignment challenge — LessWrong LessWrong (AI Alignment Forum crosspost) · 12 July 2022 Own record

   “Nevertheless, my overall impression is that most of the new people coming into alignment research end up pursuing research that seems doomed to me, not just because they're unlikely to succeed at their stated research goals, but because their stated research goals have little overlap with what seem to me to be the tricky bits.”

2. E2 On how various plans miss the hard bits of the alignment challenge — LessWrong LessWrong (AI Alignment Forum crosspost) · 12 July 2022 Own record

   “a bunch of plans like Owen's (that seem to me to just completely miss the problem),”

3. E3 Most People Start With The Same Few Bad Ideas — LessWrong LessWrong (AI Alignment Forum crosspost) · 9 September 2022 Own record

   “It turns out that most newcomers to alignment have the *same* few Bad Ideas.”

E1–E3 show LessWrong posts from before 2023-04-06 about newcomers to alignment missing the hard parts or sharing the same bad ideas. None of the evidence shows who wrote these posts, so it does not establish that Nate Soares wrote one, which the criteria require.