Ed/Tech must-reads

Is Zoom planning on building robot clones of us all? From various sources

Perhaps I’m a little flippant but a story about some legitimate concerns and some exaggerated ones about Zoom’s use of A.I. flared up last week on socials and it is worth sensechecking.

As many big online companies do, Zoom recently updated their Terms of Service - the excessively long contracts that we all agree to without reading them in order to get on with our days. Online tech blog Stack Diary pored over the fine print and noted what it felt to be shocking abuses of user data in sections 10.2 and 10.4 and breathlessly rushed to tell everyone about it.

Let’s deal with the easy one first, section 10.4. At first glance, this might appear alarming, telling us that Zoom claims a:

I saw many comments online along the lines of how this text shows that Zoom are trying to steal our souls or something. I am not a lawyer (please feel free to chime in in the comments if you are) but this looks a lot like the boilerplate legalese in virtually every TOS for any online service. Its purpose, in a nutshell, is that we allow the company to host our words/images/audio in an online platform and we won’t sue them for copyright violations if they do so. So when I see people getting antsy about this kind of language, which, admittedly is confronting, I have to wonder about their understanding of all the issues.

Section 10.2 though, was and probably still is somewhat murkier. Particularly in an age where harvesting our data to feed to A.I. robots (LLMs, whatever) for fun and profit is actually a real concern. The Hollywood writers’ and actors’ strikes are just the starting point there of a wildly messy and complex can of worms. (Well, yes, and also visual artists have been sounding the alarm for even longer).

Section 10.2

The problem here is the stark ambiguity of “similar content or data” - which, arguably could include our voices, images, chat and screensharing.

Like most of techworld, Zoom is getting into the A.I. business to add rich functionality to their tool. Automatic captioning has been with us for a few years (and continues to improve) and auto-translation is also available to the privileged few. I must admit, I am still in two minds about the rise of automated summaries of meetings - I wonder what is missing and I am concerned that it makes me too lazy to check but the convenience of the detail is hard to deny.

Recognising an emerging PR disaster, Zoom got on the front foot quickly, publishing their own frustratingly vague (but at least timely) response that assured users that data wasn’t being used without consent. The problem here being whether agreeing to the TOS implies consent or not. (This hasn’t been resolved, to my satisfaction at least).

Further hullabaloo emerged along these lines and Zoom expanded their response stating explicitly that:

The Stack Diary blog includes a decent timeline of how they saw the story evolve. There were a number of hasty discussions in the EdTech background in many institutions and for now the story appears to have receded into the background.

More than anything, I am reminded of this (slightly sweary) scene at the end of the Coen’s Brothers’ underrated film Burn After Reading.