Reading Audrey Watters’ 2nd Breakfast post today, stimulated a comment I wanted to make: https://2ndbreakfast.audreywatters.com/agents-versus-agency/.
But I am undergoing a major cancellation of subscriptions (I found I have been spending over £300 per year on various things over and above my internet subs and hardware maintenance). This has to stop! So, while offering comments on posts like Audrey’s is in part evidence of how good and therefore stimulating they are … BUT it’s too often a paid-for feature. A barrier to dialogue.
That’s not a gripe, but it is sad that so much good writing sits behind pay-walls that often block reading and commenting. Where once the WWW was a wide open sea of dialogue it has become a street market.
I understand that moderation is essential (way too many nutters out there) and time consuming (you write 500 words and get 5000 back in comments) (and maybe moderating comments is a clear use-case for some form of AI – n’est-ce pas?) BUT I am trying to cut down my expenses. The subscription model is a pain and a barrier.
Anyways, here’s the comment I wanted to write (I have emailed it to the 2nd Breakfast address, but who knows who sees it, and no-one can respond):
I wanted to respond to your post by reminding us that Chomsky also critiqued Skinner’s views on language acquisition because his model was basically a mechanical one that could not account for some crucial features of language acquisition in children. Chomsky’s critique was, as you say, decisive. However, I think it is an interesting observation that behaviourist language acquisition has been, in some ways, vindicated by the appearance of LLMs. That doesn’t make it correct as a model of how biological brains and minds do it, but ‘computational behaviourism’ seems to ‘work’ in the field of language acquisition. However, there’s nothing cognitive, nothing ‘thoughtful’ in the way that Big Data methods appear to work.
(But I realise that is what you are saying in your essay – I just wanted to repeat it!)
UPDATE: Audrey posted link to the essay on BSky.app, so I added an edited version of the above as a reply: https://bsky.app/profile/audreywatters.bsky.social/post/3ljzcyedjms2v