Conversation

Notices

1. Embed this notice
   Baldur Bjarnason (baldur@toot.cafe)'s status on Friday, 27-Dec-2024 20:15:43 JST Baldur Bjarnason

   “Does current AI represent a dead end? | BCS”

   https://www.bcs.org/articles-opinion-and-research/does-current-ai-represent-a-dead-end/

   > From the perspective of software engineering, current AI systems are unmanageable, and as a consequence their use in serious contexts is irresponsible.

   Been saying this for a long long while and I highly doubt anybody who is still using these systems is ever going to change their mind, but you gotta try.

   Also, I’ll bet the original title of the piece was way more forceful and was watered down into a question

   • Embed this notice
     Rich Felker (dalias@hachyderm.io)'s status on Friday, 27-Dec-2024 20:15:43 JST Rich Felker

     in reply to

     @baldur Absolutely and anyone who understood 5% of the subject matter could have told them this 2-3 years ago. Bet media preferred listening to hype peddlers who were necessarily unreliable sources rather than folks with domain knowledge & no conflicts of interest.

   • Embed this notice
     Zimmie (bob_zim@infosec.exchange)'s status on Friday, 27-Dec-2024 23:49:29 JST Zimmie

     in reply to

     • Rich Felker

     @dalias @baldur We’ve known what is currently sold as “AI” is a dead end since at least 1956 with Chomsky’s paper Three Models for the Description of Language.

     I can’t imagine how frustrating it must be for him having published on the topic for twice as long as a lot of the proponents of LLMs as AI have been alive.

   • Embed this notice
     Rich Felker (dalias@hachyderm.io)'s status on Friday, 27-Dec-2024 23:50:35 JST Rich Felker

     in reply to

     • Zimmie

     @bob_zim @baldur The old Chomsky concept of AI is bogus too. Language ≠ intelligence.

   • Embed this notice
     Rich Felker (dalias@hachyderm.io)'s status on Saturday, 28-Dec-2024 00:00:27 JST Rich Felker

     in reply to

     • Zimmie

     @bob_zim @baldur Right. Statistical models and Chomsky are the opposite directions of machine language processing with their own strengths and weaknesses. But neither is intelligence because language processing isn't intelligence.

   • Embed this notice
     Zimmie (bob_zim@infosec.exchange)'s status on Saturday, 28-Dec-2024 00:00:30 JST Zimmie

     in reply to

     • Rich Felker

     @dalias @baldur Sure, but my point is LLMs are statistical grammar. They get syntax right almost all the time, but they don’t make any attempt at semantics.