> Declaration of Interest: [Authors] have financial interests in...Arbitrus.ai. As the title would suggest, the authors are making no effort to obfuscate this fact.
2,292 karma · joined December 1, 2015
> Declaration of Interest: [Authors] have financial interests in...Arbitrus.ai. As the title would suggest, the authors are making no effort to obfuscate this fact.
Why does the author feel confident that Claude won't do this?
https://docs.python.org/3.12/reference/compound_stmts.html#f...
Where do we go from here? What kind of action would be effective?
Sometimes the peer group can be remote. The only emo kid in Baghdad is a unique snowflake relative to other Baghdadis, but is following a fashion code vetted by people on Tumblr.
I guess they'd argue that the author's identity shouldn't matter and that one should just judge the writing on its own merit. The problem is that we're drowning in information. AI makes its trivial to generate authoritative-looking nonsense. Evaluation is costly. So I prefer to read something which the author has associated with their own identity (which means there's a cost to lying), or has provided some credentials to prove why their opinion matters. Of course there are exceptions (Gwern comes to mind), but I'll only trust them if they're recommended by other writers I trust.
I have some "Tintin in Vietnam" prints at home that are commonly sold in SE Asia [1]. Those street vendors will be happy to know that they're in the clear now.
[0] See https://news.ycombinator.com/item?id=39625647
[1] eg https://old.reddit.com/r/TheAdventuresofTintin/comments/c0pv...
I would hazard a guess that with better tuning, XGBoost would still have won. (The paper notes that the authors chose a suboptimal set of hyperparameters out of fear of overfitting - maybe the same logic justifies choosing a suboptimal model type...)
Which makes me think, maybe a good move for Stack Overflow (which does not allow the submission of LLM-generated answers, wisely imo) would be to add an AI agent that would suggest an answer for each question, that people could vote on. That way you can elicit human and machine answers, and still have the verification process.