←back to thread

203 points tysone | 6 comments | | HN request time: 0.826s | source | bottom
1. dekhn ◴[] No.42199382[source]
I remember Urs arguing for this at TGIF quite some time ago. He said legal costs were increasing exponentially while the value of old email was only linear, which was unsustainable.

One outcome of this was to wipe a number of ongoing scientific discussions I was having with external collaborators. I'm used to people having the last 30 years of mail on hand to be able to carry out extremely long, complex projects.

replies(3): >>42200582 #>>42210399 #>>42215259 #
2. Ferret7446 ◴[] No.42200582[source]
Good point. To add to that, I hazard a guess that the legal costs are about collecting everything for discovery, and not necessarily about them causing Google to lose legal cases.
replies(2): >>42201412 #>>42206464 #
3. lazide ◴[] No.42201412[source]
Not just collecting for discovery - but reviewing. Manually.
4. dekhn ◴[] No.42206464[source]
What's amusing is that at that very time, Google was building its Email Discovery platform (https://support.google.com/a/answer/2462365?hl=en)- basically a way for lawyers to search, inspect, and triage emails in an org to make them available. Yes, lawyers (billing hourly) trawl through hundreds to thousands of email then hand them over to the opposition. I talked to the eng building the product and they said they personally had to do some manual email classification to train the retrieval models and that they lost any interest in using email after seeing what people put it their corporate email.
5. telotortium ◴[] No.42210399[source]
Did they not have the indef label at that time? In any case, at some point your collaborators will leave institutions and thus get a new email address. What happens then?
6. okdood64 ◴[] No.42215259[source]
> He said legal costs were increasing exponentially while the value of old email was only linear

Uh this sounds cool at first when you say it, but what does it actually mean? Can't he just say legal costs of keeping that email is more than the value of it. How did he derive exponential?