When I was looking for a new perspective to learn about Next'23, I saw these four senior database analysts talking about the conference. In the new era of AIGC, we might as well use the old-fashioned human(analyst)-powered way to summarize. The following content of this article comes from the conference dialogue of "It Depends: Google Cloud Next 23 Analyst Recap Merv Adrian, Maribel Lopez, Eli Khnaser" (you can view it directly on YouTube, a link is attached at the end of the article). These four are all former Gartner/Forrester Research VPs, among whom Merv has been the lead author of the Gartner Database Magic Quadrant for many years.
1. something impressive
- Google is very active in augmented (AI-power) multi-workload management, such as the announcement of Spanner boost (Merv).
- Google realizes the "need to play nice with other cloud services" (Maribel Lopez). Cross-cloud (hybrid and multi-cloud) and cloud-neutral will be the future (author's opinion).
- Colossus is a shared storage for Spanner, Bigquery and many applications on Google, providing the possibility of data interoperability and interaction.
2. Lagging thoughts
- Google is still in reaction mode, talking about the new world in "old" terms, still talking about TPUs, VM instances, networking. (It was refreshing to see Google 25 years ago in 1998.) Where is the search/chat bar? (Eli Khnaser);
- How are real companies using AI? (Lopez); AI is “making business and technology indistinguishable” (Eli). However (and my personal opinion is), AI is not ready to help the end customer yet. Automation and efficiency are still focused on developers (or data scientists or prompt engineers) and don’t yet provide the tools that real business people need:
1. Doctors don’t care about GPU/TPU/VM, they need a summary of the patient’s medical records (AI-Summary);
2. Tax experts or financial analysts need the tool AI to learn the changes in tax forms in 2023 (AI-learning) and clean customer data (AI-data cleaning).
3. The entry point of cloud database service
- Database migration Migration (in my opinion, the most traditional, ongoing and difficult work in databases): After most of the data transfer and schema conversion, the most difficult 5% is left (this is exactly what I helped clients from four years ago manual part when migrating from Oracle to PolarDB). Duet AI may be helpful for AlloyDB (Merv mentioned the migration from Oracle to AlloyDB, and there is also a direct demo on Google keynote);
- System Training: There are 5,000 security alerts within 10 minutes, and AI needs to detect the 5-10 that need to be checked manually? (Lopez);
- Observability for an operational system: See what happened 12 hours ago and is now crashing the system. Yes, this is indeed a problem for another cloud service provider. (Btw, #Back to the Future, my favorite movie).
4. Interesting gossip
- Eli talks about maybe it's time for AS400 to be retired, and that the old language can be replaced by a newer language more easily. For example, Google demonstrated how to use Duet AI to convert CPP's database application into GO language with one click. I hope Eli is saying it too soon, I still "dream" about going back to working with PLX on a mainframe mainframe and learned a lot from them during my years in the industry. Interestingly, Merv mentioned that IBM offers tape storage for S3;
- Lopez mentions how analyzing error code 435 is another interesting way to improve service. For me, it is also deeply touched. The SQLCode -440 back then made me stay up all night twice;
- Previous product releases were all technology-centric, listing dozens of products and hundreds of functions. This is far out of date (I can attest to this from Db2/z versions)
5. Others
- Duet AI: AI-powered copolit for helping end users (Merv)
- Vertex AI: A unified AI platform, a bit like hugging face.
The above is a fusion of my personal views corresponding to Google Next'23. May not reflect the speaker's complete opinions or may be biased. Please refer to the original video: https://www.youtube.com/watch?v=uRa5Uprwqw4
Author: Ni Demai is a database product expert from Jiuzhang Math. Nine Data NineData is a new generation of cloud-native intelligent data management platform, including data replication, SQL development, data backup and data comparison and other functions. NineData adopts leading cloud native and AIGC technologies to provide an intelligent data management platform for architecture design in the cloud and AI era. As the industry's first platform that supports seamless integration of cross-cloud and local IDC, it can help customers easily complete data migration to the cloud, real-time transmission of cross-cloud data, ETL, data backup, enterprise-level intelligent SQL development, database development specifications, production changes, and sensitive Data management and other functions to make customer data use more secure and efficient. Official website address: NineData-Let everyone make good use of data and cloud-Nine Data