Do you really think open book tests are an "obvious option" when their are whole business out there that will do the online course for a student, tests, homework, powerpoints of presentations, even coaching to get around any attempts to ask the questions on their work?

Your complaint is made irrelevant with your own statement:

None of us really like writing and grading tests. We would appreciate new methods of assessment. But we need things that actually work.

Please take a moment to look for yourself at the massive cheating operations going on, and then tell me how I can make sure my students are actually doing their work.

Telling us to just come up with "something else" is the same kind of magical thinking that this site has always criticized in the the debate over encryption. Educate yourself on the issues we face and then give us some good, practical ideas we can actually use. You’ll have a receptive, and appreciative, audience.

You are just as avoidant of work as the students you’re complaining about. Demanding that IT, who by your own admission, has no ability to teach or grade a student, magically come up with a solution for you to use.

The Internet has changed the way people get and retain information. Most people with an internet-connected device don’t memorize things anymore. They just look the information up when they need to use it and forget it again the second it’s no longer useful. Further, due to the lack of hard AI to grade responses, most automated tests and assignments are simplified to a point that no actual demonstration of intelligence is required to pass it. In cases where the test is entirely made up of multiple choice answers, it’s even possible to ace the the test / assignment through pure random chance. A fact that only serves to hasten the adoption of the use-it-and-forget-it model of "learning."

This website is fond of pointing out that content moderation doesn’t work "at scale." Do you really think continuous assessment would as well?

Continuous assessment is a teaching method, Content moderation is a censorship method. Two very opposite goals, but the means of implementing either via a computer is the same: They both require a computer to be able to derive meaning from human language. Of which no computer system is capable of doing reliably. If they were capable, chances are you wouldn’t be teaching. As by modern society’s standards, the computers would do the thinking for you, and or those without a computer would be so economically disadvantaged that mere survival would be a far bigger concern.

As for open book tests, that’s just an acceptance of the new reality. People don’t remember things, they just look it up when needed. It doesn’t matter if it’s a student doing homework, a company taking a test for payment, or in the future a computer doing the thinking for humanity. They all use-it-and-forget-it to their own short-term benefit. It is a noble effort to encourage self-reliance, one that deserves praise at every opportunity, but to think that someone won’t just look something up to save the effort of remembering something on their own is severely underestimating humanity’s laziness. If you want to stop them from doing that anyway, perhaps you need to better instill in them a sense that the effort they are avoiding now should be made anyway.

The problem is mainly the result of deskilling tutors and expanding tutorials from one-on-one or small groups to classes not much smaller than secondary school. Continuous assessment is quite straightforward in the former case (and it becomes very obvious when there is a major level of cheating) but when you’ve got hundreds of students per academic and most tutors are junior postgrads or even senior undergrads with no teaching experience, you’re asking for trouble.