Yep. Also bizarre to exclude things like philanthropic revenue but not exclude any of the expenditures funded by philanthropy

Who me?

I’m just a bush accountant, but it’s embarrassing how gullible or incurious so many academics can be about this stuff. Managers have come to expect people know nothing and don’t feel confident questioning it (and are often badly informed themselves)

If your campus is full of scaffolding, pay attention to this item in the Annual Report

Another one is categorising things as ongoing or one-off expenses as is convenient for the year’s narrative. ‘Maintenance’ is a recurring expense, but there is a lot of discretion on timing. One way of hiding a surplus is a maintenance wave, which is counted as if it will keep happening every year

Plus writing down assets; plus removing research and philanthropic income, but still counting the expenditures against that income

Another is the idea that some funds are tied to specific purposes, and so asset earnings are completely irrelevant to what the institution can spend. Some cover things the uni wouldn’t otherwise pay for (e.g. prizes funded by bequests), but others cover necessities

The other side is, of course - of course! - the companies making LLMs are not doing so with the goal of more easily sharing the material that people have made freely available on the web. They are doing so with the goal of enclosing it - converting the products of free human activity to commodities.

The lesson we should be taking from LLMs is the immense social value there is in having all kinds of material - all kinds of products of human intellectual labor - freely available online. They should be reminding us of the early utopian promise of the web.

Motherfucker this is the London review of books you better take your sensitive ass back to the Atlantic

Yes in my experience it’s more that the publisher is risk averse and also doesn’t want to bother with hassle/cost of asking for rights. If only you were training an LLM

*epistemic