In a recent X thread, Simo Ryu—who created the popular LoRA adaptation method for Stable Diffusion—floats the idea that frontier AI labs could secure an edge by deliberately sidestepping researchers known for inflating their own importance during heated hiring scrambles, especially when much of the real differentiation may come down to raw compute and data rather than individual star power.
@cloneofsimo the market for that is still huge tho
it doesn't stop people from overclaiming their own relevance and labs from snapping people up in some zero-sum talent auction
and it's also a generally useful story for regulatory capture, anti-competitive moves
@suchenzang Ok If thats all the researchers are doing id be pretty pessimistic about alphas as well
Ryu’s suggestion frames hiring as a zero-sum game where skipping the loudest voices might quietly improve team signal, though the thread offers no named targets or proof that any lab has tried this filter.
Susan Zhang pushed back that the talent pool is large enough for firms to navigate overclaims without coordination, adding that such narratives sometimes double as arguments for regulatory barriers that favor incumbents.
No Digg Deeper questions have been answered for this story yet.
@cloneofsimo not for you and i to worry about, kitten
corporations will figure it out without needing to collude against individuals
@suchenzang so maybe trying hard to not hire such researchers might be source of alpha
@suchenzang so maybe trying hard to not hire such researchers might be source of alpha
@cloneofsimo the market for that is still huge tho
it doesn't stop people from overclaiming their own relevance and labs from snapping people up in some zero-sum talent auction
and it's also a generally useful story for regulatory capture, anti-competitive moves