Industry executives and experts share their predictions for 2025. Read them in this 17th annual VMblog.com series exclusive.
By Karthik Ranganathan, co-founder and
co-CEO of Yugabyte
The data industry has experienced several
major “waves” of innovation in recent years. We’ve seen cloud-native take hold,
the shift towards microservices, the transition to infrastructure as code, and
the rise of Kubernetes. All this innovation has driven the velocity of
development from monthly releases to daily releases.
Fast forward to now and the next wave, as AI
significantly increases the rate at which we build and deploy applications.
That won’t change in 2025 as new trends and consequences emerge across the
broader data and tech industry. This is something enterprises will need to
consider carefully going forward. Here are my predictions for the year ahead.
1) AI outgrows its feature roots
to follow the path of Big Data
AI is not a feature;
it’s something you use applied to a feature, similar to Big Data. Just as we
saw with the Big Data movement transitioning from “doing Big Data” to “applying
Big Data,” 2025 will be the year where we transition from “doing AI” to “applying
AI.” This means focusing on how it can drive net new business value and impact.
2) 2025 will reveal AI’s
resilience problem
Increased AI adoption
and innovation will reveal that AI has a resilience problem that must be
addressed. Application, data center, and network outages are now all-too-common
occurrences, but how do these legacy IT challenges impact an AI-powered world? AI
applications are already resource-intensive, but the reality is that AI usage
is still fairly limited within most organizations, so we’ve only scratched the
surface of what they require in the longer term. As the adoption of AI
applications and models increases, we will see the detrimental impact of
systems and tools that haven’t prioritized ultra-resilience as a core design
principle.
3) Multi-cloud repatriation will
persist
We’ve seen the trend
of enterprises moving from private to public clouds, but in 2025 we will see AI
adoption drive a wave of simultaneous multi-cloud repatriation. Rising cloud
costs, security concerns, and resource constraints caused by AI adoption are
the main drivers behind this trend – cloud repatriation will emerge as the
strategic solution for solving it.
4) The most important tech job
of 2025 – AI translator
Enterprises are
seeing the benefits of adopting AI copilots to reduce cognitive load and get
faster results. But that doesn’t mean AI will take jobs away. For copilots to
be effective, data, documents, and policies must be formatted to allow AI
agents to consume them without generating false positives. New jobs, like AI
translators, will emerge that are focused on writing or rewriting company
documents to remove ambiguity and equip AI to make decisions.
5) Distributed databases will be
crucial for achieving AI agility
As enterprises enter
the next phase of AI development and adoption, agility will be essential.
Business needs change quickly and frequently, so enterprises have to act fast
to meet customer demand and drive value. On top of that, AI workloads are
extremely resource-intensive so teams increasingly face more data silos.
In 2025 and beyond, I
expect teams to continue AI experimentation and work on addressing data silos
to improve efficiency and performance. For enterprises trying to achieve AI
agility, having a distributed database will be crucial. These databases ensure
scalability and consistency for AI/ML initiatives by eliminating silos and
creating a stable, always-available environment for data management and access.
At re:Invent 2024,
AWS announced Aurora DSQL-a PostgreSQL-compatible, multi-region, horizontally
scalable database. This validates our belief that more applications will need
distributed data capabilities in the future through PostgreSQL, and it will become
mainstream.
2025 will be a transformative year for AI.
We’ve only just started to scratch the surface of what AI can achieve for all
businesses, from healthcare and financial services to retail and beyond. To
take advantage of this pivotal year for AI, enterprises will need to ensure
strong tech and data foundations. Once established, I expect we’ll see
unprecedented innovation across all industries.
##
ABOUT THE AUTHOR
Karthik Ranganathan is co-founder and
co-CEO of Yugabyte,
the company behind YugabyteDB, the open-source, high-performance, distributed
PostgreSQL database. Karthik is a seasoned data expert and former Facebook
engineer who founded Yugabyte alongside two of his Facebook colleagues to
revolutionize distributed databases.
