GPT 时代已经结束？

https://archive.is/2024.12.07-083509/https://www.theatlantic.com/technology/archive/2024/12/openai-o1-reasoning-models/680906/

Even if you accept the claim that o1 understands, instead of mimicking, the logic that underlies its responses, the program might actually be further from general intelligence than ChatGPT. o1’s improvements are constrained to specific subjects where you can confirm whether a solution is true—like checking a proof against mathematical laws or testing computer code for bugs. There’s no objective rubric for beautiful poetry, persuasive rhetoric, or emotional empathy with which to train the model. That likely makes o1 more narrowly applicable than GPT-4o, the University of Pennsylvania’s Rao said, which even OpenAI’s blog post announcing the model hinted at, stating: “For many common cases GPT-4o will be more capable in the near term.”

所有跟帖： 

• 按照这个文的观点，实际对找工是个好事？可以是强大的辅助工具，但并不能取代真人 -oryzivore- ♂ (0 bytes) () 12/08/2024 postreply 09:31:40

• AI 做的好的：通过数学定律验证证明或测试计算机代码的错误 -oryzivore- ♂ (0 bytes) () 12/08/2024 postreply 09:32:21

• AI 做不了的：说服性修辞，情感共鸣 -oryzivore- ♂ (0 bytes) () 12/08/2024 postreply 09:33:07

• 知识性的是AI的特长，创造性的是盲点 -我是谁的谁- ♂ (0 bytes) () 12/08/2024 postreply 09:35:47

• 常识吧。人也做不到是无所不能的全才，AI也一样。一定会分细分领域才能深入。我知道金融领域最牛的AI是不公开的 -我是谁的谁- ♂ (0 bytes) () 12/08/2024 postreply 09:34:54

• 事关gen AI 或者 super intelligence 可能性。投了billions 浪费了几个核电厂的能源，如果 -oryzivore- ♂ (0 bytes) () 12/08/2024 postreply 09:39:39

• 你去把当前的DNN看一遍就知道只不过是深度学习机器，没有什么I在里面。将来的I一定来自每个领域的模型实现。不可能有 -我是谁的谁- ♂ (53 bytes) () 12/08/2024 postreply 09:43:31

• 不能突破maze rat，那。。。 -oryzivore- ♂ (0 bytes) () 12/08/2024 postreply 09:40:38

• GPT就是把AI通俗化，在社会掀起了波浪，消耗了不少能源， 把女大送上了神坛。 -土无垠- ♂ (0 bytes) () 12/08/2024 postreply 09:38:16

• 达康2.0 -我是谁的谁- ♂ (0 bytes) () 12/08/2024 postreply 09:45:49

• 是先普及，再精耕细作。跟达康还不一样。有点像先铺了好多铁轨，在通一班班高铁 -土无垠- ♂ (0 bytes) () 12/08/2024 postreply 09:50:22

• $200一个月是败招 -ClearCase- ♂ (0 bytes) () 12/08/2024 postreply 11:12:00