Keyang Zheng
@albsun.bsky.social
PhD Candidate @ Pitt | HCI | I love playing games and studying how people play games.
An interesting observation on my twitter feed and Bluesky feed. They read like I’m looking at two completely different worlds. One is filled with posts about LLM work flows and how good LLM agents are. While the other has a lot criticism towards adopting LLM agents to do everything.
Back at Pittsburgh after PyCon US, how come the weather is warmer than all of the 5 days I spent in LA…
"peekaboo as an eval function for infant learning" is a really great way to make the point that not all intelligence is expressible in words so it is highly unlikely that large language models will give rise to AGI (via @albsun) #PyConUS
I think it might be the time to start building all the gunpla models I have collected.
Keyang Zheng@albsun.bsky.social · 5mo agoPassed my PhD thesis defense just before my birthday!
Feeling weirdly exhausted and energized at the same time after the CHI deadline...
