I think especially on like sub-seasonal timescales, like multiple months out, we're going to be able to frame the statistical question with a lot more specificity and probably a much better quantification of the uncertainty.

我认为，特别是在像次季节性的时间尺度上，比如几个月后， 我们将能够更具体地构建问题， 并且能更好地量化不确定性。

Just choose more or less on a different player's stats.

只需在不同球员的数据上或多或少地进行选。

And a t-distribution is essentially, the best way to think about is it's almost engineered so it gives a better estimate of your confidence intervals and all of that when you do have a small sample size.

t分布本质上是这样设的，当样本量较小的时候， 它能更好地估你的置信区间等量。

If we're guided by color blindness, our goal should be to eliminate bias from systems that affect people's liveswherever possible, not to manufacture statistically equal outcomes by any means necessary.

如果我们以色盲理念为指导， 我们的目标应该是尽能地从影响人们生活的系除偏见，而不是不惜一切代价制造上的平等结果。

Whether it’s because people are more anonymous or perhaps more clear about their future goals, there is a statistical increase in successful partnerships that started online.

无论是因为人们在线上更加匿名，还是更明确地表达了他们的未来目标，在线上开始的成功伴侣关系在上有所增加。

So how does ScribeN scale up without resorting to just averaging everything together, which we know kills the signal, or subsampling, which is statistically problematic?

那么，ScribeN是如何实现扩展的呢？它既没有简单地对所有数据求平均（我们知道这会破坏信号），也没有采用抽样，因为抽样在上存在问题。

It has no idea that it is actually spewing out something that appears statistically, stochastically as the next logical token to show up.

它并不知道实际上自己正在喷涌而出的是一些在上、随机性地看似下一个合乎逻辑出现的符号。

Because money now formally changes hands in the marketplace, rather than informally between husband and wife in the home, today's statistics count as output things that did not get counted before.

由于金钱现在在市场上正式易手，而不是在家丈夫和妻子之间非正式地转手，因此今天的数据被视为以前未算过的产出。

Americans, empirically speaking, just like speaking empirically about Harvard students and looking at the trends and the statistics, empirically speaking, Americans are the most generous people in the world.

美国人，从实证上讲， 就像实证地谈论哈佛学生， 看趋势和数据一样，从实证上讲， 美国人是世界上最慷慨的人。