Joel R Posted December 1, 2023 Posted December 1, 2023 (edited) 7 hours ago, Matt said: I still think that a summary will rob the community of its character and voices. It’s still important to see the author next to their content IMO. My personal take is that we can wait for Google to scrape all of our content to start offering AI generated summaries on Google (which is already happening - hello Bard!), or we can start doing it ourselves. I'm sure there's a worthy existential debate about the importance of sourcing content from reputable sources or trusted people on forums, but at the end of the day, if I'm a user searching for the fastest and most accurate solution, do I want a condensed topic (because honestly that's all the topic summary really is) of 30 discrete replies that I still need to read through and interpret or do I want an actual well-written summary that's short, sweet, to the point. There's a race to convenience that LLMs are unleashing on our public content. We can either own those summaries ourselves, with links and references to other parts of our community, or we can let Google do it for us without ever giving us any traffic. Edited December 1, 2023 by Joel R Rikki 1
Mike G. Posted December 1, 2023 Posted December 1, 2023 4 hours ago, Joel R said: My personal take is that we can wait for Google to scrape all of our content to start offering AI generated summaries on Google (which is already happening - hello Bard!), or we can start doing it ourselves. I'm sure there's a worthy existential debate about the importance of sourcing content from reputable sources or trusted people on forums, but at the end of the day, if I'm a user searching for the fastest and most accurate solution, do I want a condensed topic (because honestly that's all the topic summary really is) of 30 discrete replies that I still need to read through and interpret or do I want an actual well-written summary that's short, sweet, to the point. There's a race to convenience that LLMs are unleashing on our public content. We can either own those summaries ourselves, with links and references to other parts of our community, or we can let Google do it for us without ever giving us any traffic. Hey @Joel R, I hear you on that front and I don't think there's a right or wrong answer regarding the AI-generated summaries. Maybe "Topic summary" isn't the best terminology if it's causing some confusion. The way I'm looking at the Topic Summary is as if it were a 'Cliffs Notes' version of the topic. Also, to quickly touch on the point that I bolded above - If used the way its intended, the new Helpful classifier should help users get to the most accurate solutions quickly. I think the AI Topic Summaries is absolutely worth more discussion and consideration, but it's something that needs to be done right if or when it is implemented. konon 1
Dll Posted December 2, 2023 Posted December 2, 2023 14 hours ago, Joel R said: My personal take is that we can wait for Google to scrape all of our content to start offering AI generated summaries on Google (which is already happening - hello Bard!), or we can start doing it ourselves. You can block ai scrapers to stop that happening. https://www.iubenda.com/en/help/137640-block-openai-bard-crawlers David N. 1
My Sharona Posted December 2, 2023 Posted December 2, 2023 Apologies if addressed already. Will there be an optional ability to turn this feature on (or off for that matter) within the ACP?
Marc Posted December 4, 2023 Posted December 4, 2023 On 12/2/2023 at 1:35 PM, My Sharona said: Apologies if addressed already. Will there be an optional ability to turn this feature on (or off for that matter) within the ACP? Yes 🙂 My Sharona, David N., Mike G. and 1 other 1 3
Max Posted December 27, 2023 Posted December 27, 2023 On 11/29/2023 at 4:18 PM, Matt said: The problem is that too many people want to throw AI at everything. I've seen loads of AI based 'executive summaries' on things like Helpscout, Amazon Reviews, etc that honestly do not offer much value. We are happy with using signals and a mathematical model for this feature. @Matt Finger has been discussing building a LLM for many things, so it's an area we are interested in. This is so exciting. Would love to have an AI instance that has read every one of our posts and is available for questions. Tell me what x user said about this event in 2021. Big dreams, I know. 🙂 Matt 1
Recommended Posts