Hi Alex, love your deep dives. If not already, could you also please deep dive into AI podcast creation (two-person conversational style)? Tech stack and how to replicate for inspiration please?
I’m referring to rGoogle’s NotebookLM’s recently launched 📓 AI podcast creation capability. It’s amazing how we are able to create engaging, and candid two-person (read robot) conversations based on our own local content.
I have one doubt regarding Active passive replication where I am not able to understand how this replication strategy ensures strong consistency. Can someone help me in understanding this ?
As per my understanding the intention of author is to state that in active-active mode as mentioned in diagram, both the sites process data and update local data , so there could be some discrepancies like conflicts or other deviations resulting in less consistent data across. Whereas in active-passive mode, only one consumer is allowed to process so thus having less of above mentioned problems. Just my thoughts .
Isn't it better if you'll keep some questions which the article will answer in the beginning of the articles? Like if anyone before reading it can think the answers about those questions on their own, it'll build interest for everyone to read the article and digest the content more deeply in mind.
As of now, it just feels like reading it, knowing the how and then, forgetting it later. If we will think about some solution on their own, we'll better be able to digest the article.
Hi Alex, love your deep dives. If not already, could you also please deep dive into AI podcast creation (two-person conversational style)? Tech stack and how to replicate for inspiration please?
I’m referring to rGoogle’s NotebookLM’s recently launched 📓 AI podcast creation capability. It’s amazing how we are able to create engaging, and candid two-person (read robot) conversations based on our own local content.
Here’s my scoop published today for reference-
https://open.substack.com/pub/millionai/p/create-ai-podcasts-in-seconds-without?r=297y6u&utm_medium=ios
I have one doubt regarding Active passive replication where I am not able to understand how this replication strategy ensures strong consistency. Can someone help me in understanding this ?
As per my understanding the intention of author is to state that in active-active mode as mentioned in diagram, both the sites process data and update local data , so there could be some discrepancies like conflicts or other deviations resulting in less consistent data across. Whereas in active-passive mode, only one consumer is allowed to process so thus having less of above mentioned problems. Just my thoughts .
Makes sense. This is what I am assuming as well after reading the original paper.
Hey Alex, Just a suggestion.
Isn't it better if you'll keep some questions which the article will answer in the beginning of the articles? Like if anyone before reading it can think the answers about those questions on their own, it'll build interest for everyone to read the article and digest the content more deeply in mind.
As of now, it just feels like reading it, knowing the how and then, forgetting it later. If we will think about some solution on their own, we'll better be able to digest the article.