21
1本期嘉宾：SAIR Foundation联合创始人、菲尔兹奖得主、UCLA数学教授陶哲轩（Terence Tao），与他的长期挚友、计算艺术史学者Tanya Klowden。一位深耕纯数学，一位专攻16世纪西班牙肖像画，却共同完成了论文《AI时代的数学方法与人类思维》。
【时间线】
00:06 嘉宾介绍及研究背景
01:21 论文背景与目的
03:30 AI在数学和科学中的应用
05:12 AI的信任与关系建立
06:20 AI的伦理与社会影响
06:31 论文对读者的影响与期望
07:12 AI在数学研究中的局限性
12:23 合作背景与过程
15:20 合作中的挑战与解决方案
19:03 智能的多样性与合作的重要性
24:29 建立与AI的关系
28:07 关于建议与指导的看法
【核心观点摘录】
“I think we are now having to face basic philosophical questions which we had the luxury of ignoring in the past.”
“What does it mean for a proof to be correct, to be interesting? What is the actual purpose of our discipline?”
“Humans, when they want to visit a waterfall, they will hike and look at maps, and it's a journey... Now we have tools that can fly us right to the destination. We can get our Instagram shot and go home. We technically accomplished what we wanted to do, but it somehow feels empty.”
“We have to have a much richer version of what intelligence is. We closed our paper with what we call a Copernican view of intelligence.”
“It's not a one-dimensional linear narrative that AI are either sub-human, human, or superhuman intelligence... We have to think in much wider terms than just what humans alone or what computers alone can do.”
“You should get advice from many, many sources. Not just the most successful person in your field, because that's a very biased sample. Also get advice from people who are in a similar situation to you, maybe just a couple years ahead.”
“Don't be afraid to ask the stupid questions. Don't be afraid to be wrong.”
“There's something that a collaboration can create that no single person in the collaboration could have done separately.”
【适合谁听】
- 对AI时代人类角色感到困惑的知识工作者
- 关心跨学科合作与科研本质的研究者
- 面对AI焦虑想重新定位自己的学生与教师
- 喜欢跨界思考、哲学与科技交汇话题的读者
- 所有想从“人vs机器”对立思维中跳出来的人
