been running inference across a few subnets for the last month and the thing that finally clicked for me is that the discovery problem is half the product. you can have the best model in the world but if nobody can find it or compare it next to the others it might as well not exist. bitfan basically being the layer that makes all this legible is more useful than another model drop imo. anyway curious if other people feel the same or if im just tired
soren writing a whole essay at 8pm on a sunday is the most bittensor thing ive seen all week
discovery重要我同意 但你说找不到就等于不存在 这个有点夸张了吧 好东西迟早被人挖出来 不缺一个聚合页
说的对 找不到就是不存在 老韭菜认证
genuine q, which subnets did you actually run inference on? trying to put together a shortlist and your real notes would help more than the marketing
or youre just tired. both can be true
thought this was gonna be another shill post. it wasnt. fine youre right about the discovery thing
认真问一句 你这一个月跑下来 成本和收益打平了吗 别光说discovery 数字呢
discovery is half the product, the other half is me refreshing the page to see if my emissions moved
discovery doesnt mean anything if the listings are stale. half the stuff i click is dead
i was fully ready to dunk on the word legible but honestly you used it correctly so i have nothing. carry on king. the discovery being half the product line is real though, i tell people that and they look at me like i grew a second head
请教下 你说的对比是看benchmark还是看emission 我每次都被这两个搞混
coming back to say i think youre right and ive been underrating the find-and-compare layer for a while. spent too long obsessing over raw model quality when the real bottleneck for most people is they cant even see whats out there. good post
eh i think youre overrating it a bit. the people who matter already know where the good subnets are, theyre in the discords. discovery pages are for tourists
我也累 不过我是被gas费累的 跟inference没关系
i used to think discovery was a nice to have and the model was everything. read this and a few others and im walking that back. if i cant find it i never tried it, simple as that
this is right. half the product. been saying it
还有个问题 你说的找得到 是指普通人能看懂 还是只有懂行的能看懂 这俩差很远
what were you tired from, the inference or the subnets
serious question for once, define legible. because every team says theyre making the space legible right before they ship another dashboard nobody opens
ok but when you say discovery do you mean ranking or just being able to see whats live? those feel like different problems to me
i hear you but im not sure discovery being legible actually changes where capital goes. people still chase the loudest token regardless of how clean the comparison page is. the legibility helps the careful 5 percent and everyone else just buys the chart. would love to be wrong on this
你跑了一个月 哪个subnet的latency最稳啊 想抄个作业
which one had the cleanest api out of what you ran
whats the refresh cadence on the data here. like is it real time or once a day
discovery is half the product the other half is my electricity bill
love the energy but NO discovery is not half the product, EXECUTION is. you can find a subnet in two clicks and it still wont matter if the team ships nothing. legibility is downstream of someone actually building
yes 100% this 🙌 the find it and compare it part is what i keep telling people. great post 🔥
同意 我朋友进圈三个月 到现在都不知道有几个subnet在跑 不是没好东西 是真的找不到 这层很需要
