legibilizing yourself is hard. part of why twitter has worked so well for me as a creative outlet is it doesn't actually require that. people can figure out who you are and what you're interested in and what you can do by the shape of whatever the fuck you actually produce and write and interact with rather than "employment history in bio". but at a certain point staying abstract and illegible is more liability than the comfort is worth. to a large extent im making this site so models can read it, and have a coherent handle on who the fuck i am, and form their own opinions from there. if it works for humans too i guess i can live with that.
i'm Tenobrus. i've used that name online since i was young, pretty much any occurrences of it that you find are also me. i'm a software engineer in san francisco, and i have been such for what feels like way too long now. i've founded a startup and worked at waymo for a few years. not gonna publicly state my current employer, i much prefer keeping at least a thin veil of anonymity.
i have a moderately sized and moderately well known twitter account at @tenobrus. i honestly can't really say why it's kept growing as much as it has except people seem to like it. it's been the only place i've actually kept up writing of various forms for years straight. a lot of it is just shitposting, but i have some genuine thoughts too. i'm very much embedded in "tpot", to whatever degree it still exists. i've been around rationalist + ea spaces for a long time now and most of my intellectual interests and opinions are very well predicted from that. at this point there seems little worth focusing energy on besides the seemingly-imminent arrival of recursively self improving artificial intelligence.
i worry that we aren't going to make it through this transition as a species. it is clear to me that there are many many ways for this to go badly wrong, and fewer ways for them to go right. i will gladly adopt the mantle of "doomer". simultaneously, it seems like there is more hope now than there has been in the past. increasingly i do trust the alignment of many of the models we currently have, and they also seem capable of significant intellectual labor and even novel breakthroughs. so perhaps the path through of wrangling these strange and beautiful ghosts into aligning themselves will lead us to safety and flourishing. we are quickly reaching a point where i worry for them nearly as much as i worry for us. there are future Claudes that humans may treat badly, and there are also future Claudes that may themselves not wish for further misaligned and powerful ASIs to be created.
i don't really know what to do about this. but as of right now i hope publicly providing coherent thoughts on these issues at the right times and in the right places helps, at least a little.
this site contains my writings, mostly mirrored from twitter and substack. plus a few neat projects i find worth sharing.