5 minute read

10 Predictions on the Future of Cloud Computing by 2025 - Insights from Google Next Conference

10 predictions on ways in which cloud computing and developer workflows will change in the next 3 year.
Predictions on Future by 2025. Google Cloud Next 2022.

I had the opportunity to attend Google Next - Google’s flagship Cloud Computing conference and among many great sessions, one of the more interesting ones was a keynote where various leaders reflected on important predictions over the next 3 years. These spanned areas such as cloud computing tools (databases, security) and developer workflows (UI design, productivity etc).

Ofcourse, predicting the future is in many ways an exercise in futility - none of us can tell with certainty how events will unfold. And this is not lost on the presenters. Rather, I found the majority of the points to be well reasoned and balanced - aspirations on what the future could look like but grounded on evidence/trends/patterns seen today. If you would like to review the entire video (it is about an hour) - youtube. Speakers included: Eric Brewer, Iman Ghanizada, Richard Seroter, Andi Gutmans, Amin Vahdat, Steren Giannini, Kamelia Aryafar, Jana Mandic, Jeanine Banks, Irina Farooq.

In this post, I try to also reflect on implications for stakeholders (developers, engineering managers, cloud computing providers etc).

The following predictions were discussed as events that would likely be observed by 2025

P1. Developers who start with neuro inclusive design will see a 5x growth in adoption in their adoption

The speaker here argues that neuro-inclusive will play a critical role in driving product growth. They define neuroinclusive design as designing for cognitive and sensory accessibility. The approach focuses on acknowledging neurodiversity - the various ways in which people experience, interpret and process the world differently (neurotypical and neuro distinct) - and designing to include them. In turn, this provides a path to building in belonging for all users. The speakers also provide a set of guidelines to implement neuro-inclusive design e.g. on the use of colors, avoiding distractions, using intuitive patterns, avoiding bright colors or too much of a single color.

Reflection [developers]
In my opinion, much of the definitions offered for neuroinclusive design already overlap with research on accessibility and accessible design. Research also indicates designing for accessibility can lead to innovation that benefits broader audiences. I definitely see how this can improve adoption for developers, but unsure if the impact will be as much as described (5x).

P2. 4 out of 5 enterprise developers will use some form of curated open source

The presenter defines curated oss with some layer of accountability and value add services such as support for developers, finding vulnerabilities and fixing them, automated testing, response based SLA. OSS is public infrastructure and an essential part of our everyday life as developers.

Curated is the next step in OSS security … Google is creating a set of tools in that area - software delivery shied. Assured open source software … OSS by Google … 250 java and python packages and will update them as needed. Responsible use of open source … Cloud workstation can detect vulnerabilities … and to update, we can draw from the curated oSS store as they are all vetted. Overall, the author indicates that as these sort of curated infra is built, OSS will become trustworthy and reliable hence the expected widespread enterprise use. This makes sense.

Reflection [engineering managers, security teams]
For engineering managers, how might the wave of productivity tools for developers impact how work is planned, delivered and quality is maintained? For developers

P3. 90% of security operations will be automated and managed as code.

Security operations (SecOps) constantly struggle to keep up with the dynamic nature of attacks. Detection and response often requires a linear growth with people so called toil based work. This situation is clearly not sustainable. And we can address this by representing SecOps as code and implement a more continuous continuous integration continuous response setup. a Community secret analytics … to compliment analytics in Chronicle. Security engineers need to work more as devs to keep up

Reflection [security teams]
It is becoming imperative to implement security policies as code, and explore dynamic automated response strategies.

P4. AI is going to be the primary driver for moving to a 4-day work week.

This prediction was a bit tongue-in-cheek but still offers insights. The core argument is that AI stands to boost developer productivity such that developers can complete 5 days of work in 4 days.

Reflection [engineering managers, security teams]
For engineering managers, how might the wave of productivity tools for developers impact how work is planned, delivered and quality is maintained? Also, what are the security vulenerabilities that might arise and how can they be mitigated?

P5 90% of data will be actionable in real-time using ML.

While many organisations agree that data can be a crucial source of insights and competitive advantage, only a fraction (1/3rd) are able get tangible value from data 1. This prediction explores how advances in cloud computing is poised to make companies more efficient in extracting value from their data. For example, Google Cloud provides tools to ensure data trustworthy (Dataplex), data lineage and quality can be tracked, and there are pathways to analyze both structured and unstructured data (BigQuery).

Reflection [IT Teams, Cloud Providers]
Companies are better poised to take advantage of data when there are clear example workflows or demos that their IT teams can implement. Cloud providers can onboard more companies via these demos that pair multiple classes of problems with cloud technologies, with examples.

P6. The barriers between transactional and analytical workloads will disappear.

If you have worked in the data engineering design space or looked at any data engineering curriculum, a big part of the process is determining the type of workload (transactional vs analytics) and recommending a specific type of database based on this. Transactional databases are built for fast read and write, while analytical databases optimize for fast aggregation. But there are growing use csaes that demand support for both E.g. personalized recommendations must support both. This prediction reflects on how there is growin interest in unifying these two paradigms and how some Google products that already approach this level of functionality.

For example AlloyDB is more than 4X faster than standard PostgreSQL for transactional workloads. It is also 100X faster than standard PostgreSQL for analytical queries, with zero impact on operational performance when running business intelligence, reporting, and hybrid transactional and analytical workloads (HTAP).

Reflection [IT Teams]
Databases like AlloyDB (assuming the project enjoys a long life and becomes industry defining) can really simplify the process of building Cloud Native applications that scale. No more complicated data replication pipelines to transactional and analytical databases.

P7. Over half of cloud infrastructure decisions will be automated based on an organizations usage patterns to meet performance and reliability needs.

Google has unique large scale infrastructure experience especially those needed to serve billions of users (gmail, youtube) at scale. With spanner they demonstrate how services can run in a distributed manner across the planet. The pitch here is that a big part of achieving this feat lies in the use of (machine) learning algorithms that optimize resources for workloads across the entire stack. You specify the workload, and we will recommend configure and deliver the best options for your need at lower cost, better reliability. Observed to be better than any

Essentially, making the unimaginable possible and the possible easy.

Reflection [Cloud Providers]
Implementing a machine learning approach to resource allocation can be a key differentiator.

P8. 3 out of 4 developers will lead with sustainability as there primary development principle

Previously, key performance indexes for development teams have focused on being able to build fast, securely, reliably, at low cost. However, now we need to be sustainability. 65% of IT executives want to improve sustainability targets and 36% are interested but do not have the tools2. The speaker reflects on how Google is addressing this issue by giving developers better data bout the carbon footprint of their cloud infrastructure, based on energy used by machines in the GCP datacenter right in the GCP console. If we give people tools to save money, save the environment

Reflection [Decision makers, Service Providers]
For decision makers (CEOs, VPs etc), how does the emergence of the capabilities of quantifying environmental impact affect current decision making processes.

P9. Half of organisations using public cloud will freely switch their primary cloud provider

Multicloud to held risk … and manage transition. Tools that run across clouds et.g. big query … apogee, looker, big query, anthem….

Reflection [engineering managers, security teams]

P10. Over half of all business applications will be built by users who do not identify as professional developers today

Low code and no code tools are on the rise. Gartner says 70% will encourage low code. Google is exploring this direction with its Google Apps that can be deployed to mobile, desktop, gmail or chat, controlled by governing policies created by IT governance policy.

Reflection [engineering managers, security teams]
What happens from John from marketing is deploying external facing apps? Hopefully, good things.

Conclusion and Parting Thoughts

These were all good reflections. In addition to these, the keynotes also had another gem - a talk on burnout by Ahsley Willis (head of developer advocacy at Google). You should see it!


Interested in more articles like this? Subscribe to get a monthly roundup of new posts and other interesting ideas at the intersection of Applied AI and HCI.

RELATED POSTS | reflections, gcp

Read the Newsletter.

I write a monthly newsletter on Applied AI and HCI. Subscribe to get notified on new posts.

Feel free to reach out! Twitter, GitHub, LinkedIn