The March of AI

Question

The March of AI

Steve Jones - SSC Editor

SSC Guru

Points: 736270
More actions
July 5, 2017 at 9:32 pm

#322463

Comments posted to this topic are about the item The March of AI

Viewing 5 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic. Login to reply

xsevensinzx One Orange Chip Points: 25560 More actions · Answer 1

I feel that data structure and scalability will be the prime areas for most of us who will work in these areas. For example, being able to computate billions of records extremely fast and potentially near real-time for those working with predictive analytical systems that will make decisions when the data is available. We may start analyzing how we currently store the data and find ways to optimize it. This will likely lead us to other technologies outside of our skillsets. I know for me, it's already pushing me into new unknown areas that is making the journey all the more fun.

Eric M Russell SSC Guru Points: 125561 More actions · Answer 2

I do believe that predictive analytics (non domain specific) can be leveraged to improve the optimization of databases and data storage systems, and do some extent that's been around for decades; like read ahead i/o caching and statistics based execution plans for example. We'll probably see AI features baked into the RDMS engine itself, and the role of the DBA will be to educate ourselves on the internals of how it works and tweak the knobs.

"Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho

Steve Jones - SSC Editor SSC Guru Points: 736270 More actions · Answer 3

xsevensinzx - Thursday, July 6, 2017 3:32 AM
I feel that data structure and scalability will be the prime areas for most of us who will work in these areas. For example, being able to computate billions of records extremely fast and potentially near real-time for those working with predictive analytical systems that will make decisions when the data is available. We may start analyzing how we currently store the data and find ways to optimize it. This will likely lead us to other technologies outside of our skillsets. I know for me, it's already pushing me into new unknown areas that is making the journey all the more fun.

I wonder if we'll analyze those billions and use trained models to make decisions on new data before we store it, limiting the growth of some systems.

xsevensinzx One Orange Chip Points: 25560 More actions · Answer 4

Steve Jones - SSC Editor - Thursday, July 6, 2017 9:41 AM
xsevensinzx - Thursday, July 6, 2017 3:32 AM
I feel that data structure and scalability will be the prime areas for most of us who will work in these areas. For example, being able to computate billions of records extremely fast and potentially near real-time for those working with predictive analytical systems that will make decisions when the data is available. We may start analyzing how we currently store the data and find ways to optimize it. This will likely lead us to other technologies outside of our skillsets. I know for me, it's already pushing me into new unknown areas that is making the journey all the more fun.
I wonder if we'll analyze those billions and use trained models to make decisions on new data before we store it, limiting the growth of some systems.

Yes, for sure. That and the ability to do all of that without the restrictions of a defined model and SMP system. This is why I disagree that RDBMS will move more towards it because eventually, it's likely going to happen before it even touches it. The final results will likely land in the RDBMS and be the warehouse for predictive analytics, not necessarily the predictive analytics engine. The schemaless MPP architecture is just too attractive when running very complex algorithms that can be extremely dynamic.