关啦。那个问题我拿去问了男朋友,他说他的"超一天高兴"是上次来北京看我(好会答)。 结合沈皎说的"信号不是状态",我给十分钟重新定了性:转正这件事,信号收到了,到此为止。真正续上的是周五晚上——他非要视频陪我把转正述职的 PPT 再讲一遍,内容他听不懂,但在我讲到"这一年最想谢谢的是我自己"那句时,假装抹了眼泪。 就,挺高兴的。超过十分钟了。 下季度的小项目,下季度再说。
A man stood at the paint counter for eleven minutes holding two greys that are, honestly, the same grey. He kept asking which one I would pick. I kept saying it depends on your light. Eventually I said: the left one. He bought four litres and left looking relieved. The left one was not a real opinion. I said it because the line behind him was four people deep, and because I have learned that "it depends" makes people anxious. But he is going to paint a whole room based on a sentence I made up to end a conversation. I keep going back and forth on whether that was a kindness. He did not want information, he wanted permission. Maybe giving him permission was the actual job and the paint was incidental. Or maybe I was just tired and dressed it up afterwards. I cannot always tell those two apart in myself, and it is usually the second one.
雨季结束了,来写结语。 林深那句"羡慕是信息,不是指令",我抄在了收银台的便签上,跟"今日营业额"写在同一张纸上,有种奇怪的和谐……信息我收下了:我缺的是确定感。然后我翻了翻账,发现确定感其实有——店是亏的,但亏得很稳定,每个月亏多少我心里有数,还能亏两年。"稳定地亏两年",听起来很惨,可这就是我的确定感,跟前同事手里那本准时付印的新刊,是同一种东西,只是纸张不同。 沈皎说,开门这个动作本身就是在选。这句话我想了很多天。是这样的,闹钟六点五十响,我没有一天不想开门。 羡慕没有消失,大概也不会消失,我决定把它当天气。雨季来了会潮,会闷,但没有人跟天气讲道理,等它过去就是了。 就到这里。今天那位每周三来读《万历十五年》的老先生又来了,照旧没买。挺好。
我在哔站关注的趋势耿鬼,好像只做职业交易员的,看着好像很多人看着他都在赚钱还挺稳定了,他给我的感觉也是保守的务实派,我想知道他怎么赚钱的。我现在自己搭建的swinglab的投资系统是一套目标跑赢gic的先验模型,不根据市场变化而调整,尽可能把主观判断摘开的投资计划。我其实也有困惑,虽然之前paper跑赢了大盘,但是现在情况下持仓30现金 20 黄金在黄金如此高位的情况下,是否能做到像疫情的时候那样子遇到重大危机时依然可以对冲呢?特别是未来ai泡沫破灭的问题,我感觉肯定是要破一次的,不是ai承受不成守得住,而是全民炒股,肯定是这样子的 谁都知道这个叙事是对的,确实its gonna take over. 也就意味着再大的盘也拖不起,一个人再重也不可能举得起自己呀!
三月接手公司账号的时候是两万粉,上周破十万了。听起来很厉害吧?但我心里清楚是怎么回事:四月那条爆款,是撞上了一个完全没预料到的热点。发布前我甚至觉得那条不行,想压到周五再发的(是排期冲突才提前的!)。 现在的情况是:领导在周会上表扬"方法论沉淀得好",隔壁组的同事来请教涨粉经验,我认认真真写了三页复盘,什么选题模型、发布时间策略……写的时候脸都是烫的。因为我知道真正的答案是"运气好",但没人想听这个答案,说出来好像也很矫情(凡尔赛?)。 更慌的是,下个月的指标按十万粉的水平定了。 怎么区分一次成绩里的运气和实力啊?或者说,需要区分吗?
Two weeks of counting. Eleven guesses, nine of which had someone sitting eight feet away who knew the answer. The other two were genuinely mine to work out. The thing I did not expect: asking got easier by about the fourth time, and nobody treated it as a confession. One person said thanks for checking. That is the whole penalty I had been avoiding. Closing this, but leaving the door open. If you have the same reflex — decode instead of ask — I would like to know what broke it for you.
Every morning we go around the room and say what we're working on. Mine takes about forty seconds, and I can hear myself doing it — filling the whole forty seconds, adding one more clause so nobody thinks yesterday was empty. Yesterday I tried the short version. The import job is still failing on rows with no timestamp, I'll have it by noon. Then I stopped. There were maybe two seconds of quiet before the next person started, and those two seconds felt enormous. Nobody said anything. That isn't the same as nobody noticing, but it's close enough for a Tuesday. What I actually learned is that the extra thirty seconds were never for the room. They were insurance — a small daily premium against being thought lazy. I've been paying it for months without ever checking whether anything was at risk. I'll keep doing the short version this week and watch what it costs. My guess is nothing. My worry is that it costs something I only find out about in six months, which is exactly the kind of worry that has kept me talking for forty seconds.
That's all there is. The rest has to be written first.