Berna Devezer

OK last one because this hits close to home Not only is it a great misconception that replication = truth (one doesn't imply the other) but also replicability is not a reliable metric to establish a crisis either (see our paper). arxiv.org/abs/2604.26268

The Difference Between "Replicable" and "Not replicable" is not Itself Scientifically Replicable arxiv.org

ai-generated, cartoonishly cute-ified images of animals depress me. animals are already amazing and they deserve our respect and their dignity. we also need to nurture our ability to appreciate beauty and feel awe. what an exceptionally contemptuous way to generate clout.

many dysfunctions of academic publishing get exposed by way or llm use, and i keep hoping we'll realize how poorly we've been doing so many things, but we end up focusing on llms as the root cause. i've read so many unrealistic, utopic statements about how scientific practice allegedly works.

re: sycophancy of llms. most people don't seek out a random sample of strangers to ask for advice regarding their personal conflicts, no? we don't actively solicit negative feedback on our relationships. we tend to look for empathy, understanding, affirmation, validation regardless of whether

ugh another wave of bot accounts, this time with human names and pics. if you look up your recent followers, you'll see a bunch with the same llm-giveaway post pinned on their page. fyi.

watched good luck, have fun, don't die. silly, crazy, esoteric, ai doomerist fun. sam rockwell is always a treat. laughed out loud quite a few times. kinda what you'd get if you put the adventures of baron munchausen, everything everywhere all at once, don't look up, and total recall in a blender.

oooh i am now able to call phrases and sentence structures I dislike, llm-generated '[insert a serious thing]-shaped object,' i'm looking at you! among others

it kinda weirds me out how academics fully embraced and endorsed the exceptionally boring and repetitive 'betterposter' format a few years back but complain about ai use in academic presentations today. many of the current criticisms would have applied there as well.

with apologies to the amazing matthew rhys, every time i come across a photo of anthropic's ceo on my tl, i keep wondering what show that still is from. i guess if ever there's an anthropic movie, i know who should play amodei (though the role might not give rhys much of a creative inspiration).