Posts

Comments

Comment by Taisia Terumi (taisia-terumi) on Sam Altman fired from OpenAI · 2023-11-18T01:21:35.880Z · LW · GW

Yeah, I also think this is very unlikely. Just had to point out the possibility for completeness sake.

In other news, someone on Twitter (a leaker? not sure) said that there probably will be more firings and that this is a struggle of for-profit vs non-profit sides of the company, with Sama representing the for-profit side.

Comment by Taisia Terumi (taisia-terumi) on Sam Altman fired from OpenAI · 2023-11-17T21:29:18.129Z · LW · GW

Aside from obvious questions on how it will impact the alignment approach of OpenAI and whether or not it is a factional war of some sort, I really hope this has nothing to do with Sama's sister. Both options—"she is wrong but something convinced the OpenAI leadership that's she's right" and "she is actually right and finally gathered some proof of her claims"—are very bad. ...On the other hand, as cynical and grim as that is, sexual harassment probably won't spell a disaster down the line, unlike a power struggle among the tops of an AGI-pursuing company.

Comment by Taisia Terumi (taisia-terumi) on dalle2 comments · 2022-04-27T12:25:08.438Z · LW · GW

As in a picture book, it offers simple sentences in gigantic letters, one or two per page, each one nestled between huge comforting expanses of blank space.

There's no blank space, though. In between the sentences are interactive examples. Or is this a joke I'm not getting?

Comment by Taisia Terumi (taisia-terumi) on D&D.Sci II Evaluation and Ruleset · 2021-08-13T09:55:21.899Z · LW · GW

This challenge was intended primarily as a horrible, unfair trap for those inclined to approach problems by throwing conventional ML algorithms at relevant datasets without doing manual data exploration. 

I kinda noticed that when I tried to apply a Random Forest which, in my experience, almost no one tries to counter in their toy problems and encountered some really bad results (I couldn't even get enough mana if tried to go with its predictions no matter how I handled the data).

Soooo I made a simple feed-forward NN and returned home with ~20 gold according to the interactive web shopping. Using suspiciously powerful black boxes in problems obviously asking you to stop and think is also horrible and unfair, yet here I am.

To be honest, I did explore the data, making a correlation matrix between one-hot-encoded columns and readings/mana, and noticed some of the patterns you describe in the post. But they seemed complex, and I was sleepy, and so I went ahead and hammered in some nails with a microscope.

Very cool challenge idea overall, I greatly enjoyed doing it wrong. Thank you for sharing this on ACX. I'm a long-time lurker on LW, but this made me sign up.