OpenAI ingests petabytes of logs every day. 90% of that is me generating dog-themed song parodies. They agree; totally worth it* https://dub.sh/0L46XR3 * OpenAI has not officially commented on whether generating dog-themed song-parodies is, indeed, "worth it" (but a source tells me they agree**) ** the source is me
Lionel Palacin
3mo
I thought our observability platform was big! https://clickhouse.com/blog/scaling-observability-beyond-100pb-wide-events-replacing-otel
Mark Needham
3mo
the other 10% is from me generating Studio Ghibli images :D
Moshe Blumberg
3mo
I think it's cats as well!
Daniel Svonava
3mo
On the topic of "worth it" - do y'all have a way to quantify the business value of OLAP queries once they finally happen, to back-justify all the collection and retention and indexing of the underlying data? Is there a concept of data LTV that could help companies better model and understand this upfront investment?
🤣
Product at ClickHouse
3moI asked OpenAI what to comment on this blog, thereby increasing their logs ingest volume even futher 😅