The AI Loop: Will AI End Up Summarizing Its Own Work?
Lately, generative AI has been making its mark in many areas.
 |
|---|
| What AI thinks cooperation with it looks like. Image created in cooperation with MS Copilot |
In the world of personal productivity, especially working with texts is affected as these AI models are incredibly good at handling text.
Many productivity-enthusiasts have incorporated generative AI in their workflows: do you have a lot of information to dig through? Let generative AI summarize it for you.
But generative AI isn’t just great at summarizing text - it can also create entirely new content from just a small amount of information. Give it a bunch of bullet points and within seconds you get a well crafted text.
But there’s a catch.
If more and more people are using generative AI for text processing, won’t we stumble into a future where AI will communicate with itself? Let me elaborate on that with an example:
- Alice has a bunch of notes and wants to craft a document. She wants the document to include background information and clear definitions, but she doesn’t have much time to put it all together. So Alice decides to use generative AI to aid her. Within minutes she has crafted a (lengthy) document ready to be distributed to her colleagues.
- Bill gets the document just ten minutes before his next meeting starts. As he doesn’t have much time he uses generative AI to summarize the document. Bill ends up with a hand full of bullet-points and can review them quickly before attending the meeting.
Two people, two different needs: creating vs. consuming information. Here, generative AI is practically talking to itself:
- Notes
- Create a (longer) text
- Summarize to a short text → notes
Wouldn’t it have been more efficient if Alice had just condensed her notes into a few bullet points and shared those instead of a long document?