Interesting article, what's with the weird sucking up to Larry Ellison at the end though? It was so weird I thought it was sarcasm but it seemed like it might be genuine?
> Anything loosely connected to AI + LLMs received the bulk of the attention (rightly so, as it is a new chapter in computing)
Beyond LLMs and vector search, the scope for applying AI / machine learning _within_ databases is enormous: join planning, learned indexes, compression, workload prediction, configuration tuning etc.
Given the current pace of advances perhaps a dedicated "AI in Databases in 2024" review will be on the cards this time next year.
Agreed this is an underexplored topic, which is funny because a few years ago prior to LLMs, if you asked someone how they'd apply ML to a database, you'd get an answer more like this than about any vector indexing or LLM shenanigans.
I actually just read a paper where some new file format achieved far faster and smaller writes than the Parquet used in all of these modern big datalakehouses, and it was done by guessing like 20 different simple compressions on a sample to figure out what will work best for each chunk of data. Sounds like manual feature detection to me!
DuckPGQ is a super interesting project, since it would mean we would now have a good option for OLAP + graph queries in the same package. But afaict at the moment is a research project... I wonder if DuckDB has it in its roadmap to actually ship this feature.
A bit random but I wonder if anybody knows what happened to Tigris? It looked like a pretty promising product [1], then one day all activity just halted without any sort of announcement.
18 comments
[ 2.7 ms ] story [ 58.3 ms ] threadIt is genuine. Larry and I go way back.
Source: https://youtu.be/_UGUDVeSQSI?t=872
Beyond LLMs and vector search, the scope for applying AI / machine learning _within_ databases is enormous: join planning, learned indexes, compression, workload prediction, configuration tuning etc.
Given the current pace of advances perhaps a dedicated "AI in Databases in 2024" review will be on the cards this time next year.
I'll have to disagree about database customers becoming more savvy; cloud has brought in a whole new crackpipe for managers to smoke.
--
1: https://hn.algolia.com/?q=tigrisdata , a FoundationDB-based document DB, alternative to Mongo and DynamoDB.
Let Riak go...