and before anyone comes at me as a luddite or whatever, I work in machine learning. I work on DNNs and deep-learning for domain specific tasks for which they are well-suited and act as a useful adjunct (classification, supervised control systems, etc.). There’s many places they aren’t the right tool! LLMs and GenAI are uniquely corrosive, IMO because of their attack on creativity, their inability to state confidence on their answers, and their masquerading as the right tool for all jobs.
=> More informations about this toot | View the thread | More toots from stoneymonster@stoney.monster
text/geminiThis content has been proxied by September (ba2dc).