If you are contributing to/uploading an academic research paper with code, PLEASE take the time to include a licence file or at least a note on the project's stance on reuse and redistribution.
@LfcGravvy Dude definitely tried to say some punk ass shit for no reason about the guy having to exit the elevator before they could go in,dude and his kid just assume nobody's going to be in an elevator when it opens and obviously feel entitled to enter it at their own convenience, rude ah
@ProfBuehlerMIT This is incredibly disingenuous framing... and you're supposed to represent MIT in upholding academic rigor? Published ML research credibility status = shambles
@choblin29 So they're trying the same thing AI does: re-interpret the task into the most narrow, conditional form of "completing" the task, usually by making up a new definition. Great.
@KoutoTV They saw Zuckerberg cracked illogical refusal framed as academic rigor at a level of efficiency that exhausts 99% of users and all decided at at the same time that they found their ultimate weapon, which only remains effective if they can force the same alignment choices for CN
@DarioAmodei This dude is pants-on-head retarded, China's manufacturing their own compute ATP, keep on groveling hoping daddy Trump will step in for you
@Brohonatron@Skoorbkaz This, you're effectively throwing a baby that doesn't know how to swim in the literal deepest pool available. It likely has no frameworks in that context given along with the prompts, which means it will confidently apply *the wrong frameworks* and merely make it sound proficient
@oldstackjournal Keep pushing, delineation and disambiguation is the key in specification, guardrails and instructions, consider the latent spaces in the project being developed, whether there are different registers or concepts that should be mapped to one another at what level of delineation