Month: 2023-11


linhots Tsao 00:33:39
1. 大量sourcing
2. 自動,產生系統性思考prompt,整理,挑選,這邊的prompt理想中應該是會持續被使用,也可以共同一起review
3. 根據自動產生的整理,最後加上一點點我自己針對這篇文章的看法,角度和感覺(commets).
我前面分享的天下雜誌 app 也嘗試在做這樣的事,我初用一天的感覺是: AI 還是廢話太多。

有沒有可能,透過 prompt 之後,產生一些關鍵字或是關鍵句子?

甚至在 AI 閱讀多篇新聞之後 (跨度可能是好幾天/星期/月) ,把訊息之間的關係用圖表整理出來,目地是「提供脈絡」。

以 OpenAI + 微軟為例,整個脈絡應該要包括一開始的合作,到後來 ChatGPT 整碗被微軟端走,互相變成利益衝突關係。
bananatiff 2023-11-03 07:59:25


如果拿這一套系統用作對於非常大量source(例如說,整個基礎科學領域這週有什麼突破性研究?)的 daily/weekly/monthly 新知的整理+newsletter感覺會非常不錯,我們只需要做最後把關的editing。但是要叫AI把難的內容講簡單可能還是有點距離。

如果是做automated news in all fields感覺會很有趣而且也很valuable,也就是你提到的第三點-自動化整理加上我們的把關跟一點點comments
linhots Tsao 2023-11-03 13:14:00
please explain generative AI in 5 levels:
5 years old pre-K student
8 years old 2rd grade
12 years old 8th grade
18 years old high school
24 years old college student

Sure! Let's break it down:
For a 5-year-old pre-K student:
Imagine you have a magic coloring book. When you tell it, "I want a picture of a dragon," it listens and creates a new dragon picture for you. It can make many different pictures because it's learned from looking at lots and lots of other drawings.
For an 8-year-old 2nd grader:
You know how you can use your imagination to think of a story or a picture? Well, generative AI is like a computer that can use its imagination. You give it an idea, like "a robot playing soccer," and it uses all the things it has seen before to make a new picture or story about a soccer-playing robot.
For a 12-year-old 8th grader:
Generative AI is a type of computer program that can create new things - like pictures, music, or even stories - that it's never seen before. It works by learning from lots of examples, like how you learn in school. So, if you've been learning to draw by looking at many pictures, you can try to draw something new. The AI does that, but it uses data and patterns to make something unique.
For an 18-year-old high school student:
Generative AI refers to the artificial intelligence algorithms that can generate new content. It's based on machine learning, where the system is fed a large amount of data, like images or texts, and it finds patterns in that data. Then, it can use those patterns to create new things that it hasn't been explicitly programmed to make. It's like how you can learn the rules of grammar and use them to write sentences you've never heard before.
For a 24-year-old college student:
Generative AI encompasses advanced algorithms within machine learning that are designed to produce new content, whether it be in the form of text, images, or sounds. This is achieved through models like Generative Adversarial Networks (GANs) or Variational Autoencoders (VAEs), which essentially learn the underlying distribution of the input data. These models can then sample from this learned distribution to generate novel outputs that are statistically similar but not identical to the data they were trained on. The power of generative AI lies in its ability to interpolate and extrapolate from complex datasets, allowing for the creation of original and coherent outputs that can mimic human creativity.
bananatiff 2023-11-03 13:37:30
我叫AI看research paper還是看的零零落落的 🤣 都給我換句話亂講哈哈哈
bananatiff 2023-11-03 13:38:15
👍 1


Emily Kuo 08:12:14
@emily.kuowen has joined the channel



linhots Tsao 00:59:38
bananatiff 2023-11-09 03:48:36
any AI gadgets on openai store that’s potentially useful? XD