from Diary 2023-09-17 The reality of collaborative work with AI and the S-curve Summary: Working with AI appears monotonous from the outside, but in reality, it senses periods of stagnation in thinking, throws the content to AI, and moves on to other tasks. Human thinking follows an S-curve, and AI can be helpful during plateaus.
The case where someone wanted to see me use AI and I said, “I don’t think it will be interesting,” I tried to work with a little metacognition, but it still won’t be interesting. Because you hardly ever see the AI screen… Work A → throw to AI → work B → copy and paste AI’s result → throw B’s content → another work… and so on. The destination can be ChatGPT, Poe, Github Copilot, or my own omni.
It’s a pain in the ass just to deliver it because I use both a PC and a phone to begin with. Now I’m walking and working on my phone.
Metaphorically, my current use of omni is “Here’s what I wrote, but read it later.” and like sending a link with a message.
Human thinking is an S-curve like this, and even if you are able to output quickly one day, the speed gradually slows down over time. There is an unpredictable time lapse with that slowdown, and then another new S curve begins. In with AI’s intellectual production, I detect that my thinking has entered a stagnant period and throw the content up to that point to AI to do something else. Then the AI will develop the content in a few minutes. I don’t wait for the results, nor do I let it notify me that the results are available. I switch over and do something else. When that work is done or “stalled,” I come back and look at the AI’s output. There is a proverb that says, “A bad idea is like taking a rest,” but with this method, there is no “phase where you don’t know whether you are thinking or resting.
(I wrote this much and tossed it to AI, and while AI was writing the following, I was writing “[I see light behind the fog of literacy.) Overcoming Stagnation of Thought” is being conducted as “Intelligent collaboration between AI and humans”. (“Overcoming Stagnation of Thought” was thrown at the AI.)
relevance
- In June, he said this about vector searches
- A system where you can get fast feedback at any time
- Vector search is much faster, but I don’t use it at all anymore.
- It beats the experience of “writing it down to the last word and handing it over to the AI.”
- On 9/1, he wrote “Difference in skin feel between vector search and RAG”.
- Afford to edit.”
- 9/9 interpreted as “Recommendation with context”
BELOW_IS_AI_GENERATED
AIとの協働作業の実態とS字カーブ
2023-09-18 00:24
digest
. Collaborative work with AI detects periods of stagnation in thinking and throws the content to AI to move on to another task. Human thinking follows an S-curve, and AI can be helpful in times of stagnation; working with AI can smooth the transition from Task A to Task B. Also, collaborative work with AI can avoid stagnant periods in thinking by sensing when thinking is stagnant and throwing content to AI to move on to another task.
Relation to Fragment
. The fragment “Intelligent Collaboration Between AI and Humans” provides a specific example of AI-human collaborative work, which is directly related to the note. Specifically, the AI and human become team members and create a new form of teamwork, which is consistent with the collaborative work with AI described in the note.
deep thinking
Collaborative work between AI and humans can avoid periods of stagnant thinking by sensing periods of stagnant thinking and throwing the content to AI to move on to another task. This is based on the characteristic of an S-shaped curve in human thinking, and demonstrates the effectiveness of collaborative work between AI and humans.
**Title of Thought “Overcoming Stagnation of Thought Through Collaborative AI/Human Work.”
extra info
TITLES: ["Intelligent Collaboration Between AI and Humans"],"🤖🔁Once Chaos is Brought","🌀Once Chaos is Brought","🤖🔁Nerf","🤖🔁Nerf Intelligent collaboration between AI and humans."]
generated: 2023-09-18 00:24
This page is auto-translated from /nishio/AIとの協働作業の実態とS字カーブ using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.