Vol.025 ChatGPT的过去、现在和未来-—学生应用及科技幻想(上) episode artwork

EPISODE · Mar 2, 2023 · 56 MIN

Vol.025 ChatGPT的过去、现在和未来-—学生应用及科技幻想(上)

from 时差矿工

本期节目我们邀请了Haiwen(目前在美国东部博士一年级在读,研究方向是computational social science/data science)来和我们讨论最近大火的Chat GPT,从原理到科幻幻想,从商业落地到未来憧憬,人工智能以一种不可阻挡的方式迅速发展。本期节目一共会分为上下两期,上期主要是讨论了什么是大语言模型以及Chat GPT的工作原理,同时也会进行一些科幻的想象。下期主要是对商业化的探讨以及畅想未来人工智能。如果对人工智能AI好奇以及想了解如何使用人工智能辅助学习研究的同学一定不能错过本期节目哦!时间戳:Chat GPT(上)1:41 大语言模型的研究方向4:21 科普Chat GPT以及大语言模型到底是什么(语言模型-大语言模型-GPT-Chat GPT)9:30 为什么GPT模型的也会有一定的出错率13:00 Chat GPT一些奇奇怪怪有意思的问题19:03 检测文本是否由Chat GPT生成的原理21:50 New Bing — paper summarize的好助手27:18 科研视角:自己读文章还是用人工智能进行文献阅读总结?31:24 用人工智能读文章—类似于5分钟看电影32:30 本科、研究生以及博士对科研的不同视角39:50 科幻视角:机器是否能产生自己的意识?能否把人类自己的意识上传到网络?51:50 可以进行机器人利他模型的训练54:57 最终还是人类进行抉择55:25 芥末:人工智能模仿的痕迹较重参考阅读:1. Elicit:https://elicit.org/2. Costructure:https://costructure.vercel.app/3. Explainpaper:https://www.explainpaper.com/4. Schick, Timo, Jane Dwivedi-Yu, Roberto Dessì, Roberta Raileanu, Maria Lomeli, Luke Zettlemoyer, Nicola Cancedda, and Thomas Scialom. 2023. “Toolformer: Language Models Can Teach Themselves to Use Tools,” February. https://doi.org/10.48550/arXiv.2302.04761.5. Jakesch, Maurice, Advait Bhat, Daniel Buschek, Lior Zalmanson, and Mor Naaman. 2023. “Co-Writing with Opinionated Language Models Affects Users’ Views.” Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, February.6. Stanford CS 324 (winter 2022) Large Language Models https://stanford-cs324.github.io/winter2022/)7. Scaling Laws:大语言模型经济学(现在训练一个表现不错的模型只要 $500k!)(全文需付费 https://yiqinfu.zhubai.love/posts/2188959734177116160)8. Watermark:https://twitter.com/tomgoldsteincs/status/16182876650064035859. Horton, John J. 2023. “Large Language Models as Simulated Economic Agents: What Can We Learn from Homo Silicus?” arXiv. http://arxiv.org/abs/2301.07543.10. Kosinski, Michal. 2023. “Theory of Mind May Have Spontaneously Emerged in Large Language Models.” arXiv. http://arxiv.org/abs/2302.02083.

Episode metadata supplied by the publisher feed · Published Mar 2, 2023

Embed this episode

Ready to play

Vol.025 ChatGPT的过去、现在和未来-—学生应用及科技幻想(上)

0:00 56:40

No transcript for this episode yet

We transcribe on demand. Request one and we'll notify you when it's ready — usually under 10 minutes.

No similar episodes found.

No similar podcasts found.

Frequently Asked Questions

How long is this episode of 时差矿工?

This episode is 56 minutes long.

When was this 时差矿工 episode published?

This episode was published on March 2, 2023.

Can I download this 时差矿工 episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.
URL copied to clipboard!