2025 State of Cloud in Review

Aaron and Brian review the Year in Cloud, hand out Cloud awards, and discuss the biggest cloud trends from 2025. Maybe a few predictions will be made as well. SHOW: 985SHOW TRANSCRIPT: The Cloudcast #985 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:SHOW NOTESCLOUD & AI NEWS OF THE MONTH - NOV 2025 (show)CLOUD & AI NEWS OF THE MONTH - OCT 2025 (show)CLOUD & AI NEWS OF THE MONTH - SEPT 2025 (show)CLOUD & AI NEWS OF THE MONTH - AUG 2025 (show)CLOUD & AI NEWS OF THE MONTH - JUL 2025 (show)CLOUD & AI NEWS OF THE MONTH - JUN 2025 (show)CLOUD & AI NEWS OF THE MONTH - MAY 2025 (show)CLOUD & AI NEWS OF THE MONTH - APR 2025 (show)CLOUD & AI NEWS OF THE MONTH - MAR 2025 (show)CLOUD & AI NEWS OF THE MONTH - FEB 2025 (show)CLOUD & AI NEWS OF THE MONTH - JAN 2025 (show)2025 CLOUD YEAR IN REVIEWCloud of the year Cloud concept of the year (non-AI)Cloud memory of the yearCloud revenues in 2025Are NeoClouds in the hyperscaler discussion? Was Cloud (circa 2007) the actual downfall of Intel? (Cloud vs Mobile choice)Rank the clouds, end of 2025 (cloud draft 2025)Do outages matter if they are only once a year? Where does AWS go next?FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

How difficult will it be to train and build an AI Agent that has expertise in a given domain? Will it happen in the next year, or 3 years or 10 years? And who will benefit in the marketplace from his evolution? SHOW: 984SHOW TRANSCRIPT: The Cloudcast #984 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[Mailtrap] Try Mailtrap for free[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.SHOW NOTESOpenAI looks to train their models to replace junior bankersWHAT WOULD BE THE STAGES OF AN EXPERT AGENT?Train it on a set of standard knowledge (e.g. Masters of Accounting, Auditing, International Tax)Train it on a set of well-defined case studies, to provide industry contextTrain it on a set of adjacent case studies and other domains (business, law, specific industries)How to train corner cases?How to train gray areas like ethics, morality, or cost-benefit analysis? Who is motivated to train these experts? What would the cost of these experts be? Can it be similar to a human, or need to be a fraction, or a premium? Is there a way to build memory (e.g. experience) without disclosing client information? Is there a way to build shareable knowledge between agents for reinforcement training/learning?FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

Dan McConnell (Senior VP Product Management at @HitachiVantara) talks about how Enterprise infrastructure is having to evolve to keep up with the data challenges of AI. SHOW: 983SHOW TRANSCRIPT: The Cloudcast #983 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwNEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS" SPONSORS:[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.[Mailtrap] Try Mailtrap for freeSHOW NOTES:Hitachi Vantara (website)Topic 1 - Welcome to the show. Tell us about your background.Topic 2 - How is AI changing the way enterprises need to think about infrastructure?Topic 3 - What makes AI workloads so different from traditional workloads? How is AI exposing the limitations of broken cloud environments?Topic 4 - Where are organizations feeling the most pressure? What are the biggest misconceptions enterprises have about preparing for AI?Topic 5 - What problems come from managing separate systems? What is the biggest driving factor toward unified data platforms?Topic 6 - How is Hitachi Vantara helping customers handle the growing infrastructure demands created by AI?FEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

The PaaS of the future probably doesn't look anything like the PaaS of the past. AI should help bring in an era where we can abstract much more than just infrastructure to align technology and business. SHOW: 982SHOW TRANSCRIPT: The Cloudcast #982 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[Mailtrap] Try Mailtrap for free[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.SHOW NOTESVercel v0Building Web Apps with just English and AI (Acquired podcast, Feb 2025)Vercel on The Cloudcast (2024)Vercel on The Cloudcast (2021)8 tools to build your own PaaS (2025)IS PAAS READY TO TAKE THE NEXT STEP? Where could PaaS evolve to now?Can new PaaS services abstract the developer, and just focus on business logic and business ideas? What languages or design patterns would be mandated? (web only, mobile only, web + mobile? )Can we template “best practices” enough to be reliable?Can we template compliances needed to handle financial transactions, customer data, etc.?Can troubleshooting become an automated service?Where was PaaS in the past? (Heroku, Google AppEngine, Cloud Foundry, Kubernetes)Language specificCloud specificAbstracting the infrastructure and securityFEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

Brian Gracely (@bgracely) and Brandon Whichard (@bwhichard, @SoftwareDefTalk) discuss the top stories in Cloud and AI from November 2025.SHOW: 981SHOW TRANSCRIPT: The Cloudcast #981 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST:  "CLOUDCAST BASICS"SPONSORS:[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.[Mailtrap] Try Mailtrap for freeSHOW NOTES:Link to November 2025 News and ArticlesFEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

As the Thanksgiving weekend passes, let's look at some technology things that we're thankful for and excited about. SHOW: 980SHOW TRANSCRIPT: The Cloudcast #980 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[Mailtrap] Try Mailtrap for free[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcastSHOW NOTESTHINGS TO BE THANKFUL FOR:Thankful for healthHappy Birthday to ChatGPTThe feature on an iPhone that remembers the code sent in an SMS messageRole playing, brain storming, scenario planning on ChatGPT/GeminiTools like Vercel that let non-programmers use AI to build apps (the future of PaaS)Tools like NotebookLM making complex stuff like LLM training, RAG patterns,  True competition in the cloud market as differentiation emergesThe beginnings of alternatives to NVIDIA in the AI HW Acceleration marketThe evolution of wearables like fitness trackers, or AirPods (better sound quality, noise cancelling, language translation, etc.) FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

Evan Kaplan (@EvanKaplan, CEO @InfluxDB) talks about Physical AI and the evolving and emerging technologies required to bring AI to physical locations and activities. SHOW: 979SHOW TRANSCRIPT: The Cloudcast #979 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS" SPONSORS:[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcast[Mailtrap] Try Mailtrap for freeSHOW NOTES:InfluxData homepageEvan on The Cloudcast #394SpaceNews article on Time Series and AI in SpaceTime Series is critical to Physical AITopic 1 - Welcome back to the show, Evan. Give everyone a brief introduction.Topic 2 - We last spoke in 2019, and our goal with that show was to give everyone an introduction to time series databases. There’s a link in the show notes for those who want to go back and get a refresher. But, if folks aren’t up to speed, give everyone a quick definition of time series and its impacts in recent yearsTopic 3 - First, we need to discuss Physical AI. What is Physical AI, and how is it different from, say, GenAI or Agentic AI? It seems that AI in the mainstream equates LLMs with AI, but that isn’t correct. We are talking about deterministic AI, not probabilistic solutions. Can you explain to everyone the difference and why it matters?Topic 4 - Why is the concept of time series so crucial to Physical AI? Additionally, you provided a great analogy comparing time series data collection to low-resolution and high-resolution images. Can you explain to everyone why this is so important?Topic 5 - Let’s talk about some use cases. How and where does this intersection of Physical AI and time series impact organizations the most? Is this specific to certain industries (robotics, aerospace, IoT, etc.) or specific collection mechanisms (telemetry, sensor data, etc.)Topic 6 - Are we shifting with AI to a state that is less reactive and more proactive with an active intelligence?Topic 7 - What kind of demands do real-time, modern workflows and data streaming place on the infrastructure? When I think of time series, I think of real-time data, which means ultra-low latency and processing near the source, among other things. FEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

Once again NVIDIA had a record earnings quarter (Q3FY26), but the strength of their on-going success will be dependent on many factors that may or may not be within their control. Let's explore those broader factors.SHOW: 978SHOW TRANSCRIPT: The Cloudcast #978 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcast[Mailtrap] Try Mailtrap for free[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.SHOW NOTES:NVIDIA Earnings (Q3FY2026 - November 2025)WHAT WILL BE THE NEW METRICS AND MILESTONES TO TRACK?Customer Revenues (e.g. CoreWeave, OpenAI)“Alternatives” Revenues (e.g. Google/TPUs, AMD, China, etc.)Customer Success Stories (%ROI, Business Differentiation, Business Acceleration)Growth of Data Centers (e.g. buildouts, zoning approvals, etc.)Electricity Buildouts (e.g. nuclear, coal, alternative, regulatory changes, municipality adoption)Accounting Deep-Dives into NVIDIA (not fraud, but days receivables, inventory buybacks, etc.)$500B in back orders (Oracle, Microsoft, OpenAI, GrokAI)FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

SHOW: 977Rohan Sathe (@rssathe, CEO @NightfallAI) discusses the rise of Shadow AI, where employees unknowingly leak sensitive corporate data through generative AI tools like ChatGPT. SHOW TRANSCRIPT: The Cloudcast #977 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS" SPONSORS:[Mailtrap] Try Mailtrap for free[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcastSHOW NOTES:Sunday Perspective touches on Shadow AINightfall websiteTopic 1 - Welcome to the show, Rohan. Give everyone a brief introduction, including your time at Uber Eats.Topic 2 - How do you define Shadow AI? We hear Shadow AI compared to Shadow IT back at the start of cloud. However, this looks different because everyone's learning curve is much smaller. For Shadow IT to happen, you had to know IT (servers, storage, etc.). Is this the correct way to think about the problem?Topic 3 - How big is the Shadow AI problem today?Topic 4 - Normally, data leaks would be discovered by traditional DLP (data loss prevention) tools. In my experience, those tools have been cumbersome and clunky, and you often face the classic trade-off between user productivity and security, as well as the need to lock down access. How has this mindset evolved in the era of AI? Topic 5 - What happens when AI-powered attacks meet AI-powered defense?Topic 6 - Let's talk about the technical architecture. How does Nightfall actually work across SaaS apps, endpoints, browsers, and AI tools?FEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

Is the current level of AI funding and investment rational or irrational? Is it possible that it’s both at the same time? Let’s look at some numbers and the thought process behind them.SHOW: 976SHOW TRANSCRIPT: The Cloudcast #976 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcast[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.SHOW NOTES:A whole bunch of AI-related statsSam Altman on BG2 podcastDO WE HAVE ANY IDEA HOW TO MEASURE THE IMPACT OF AI?How much is one model better than another (e.g. Gemini vs. CoPilot)?How much improvement should a software developer get?How much improvement should a knowledge worker get?How much cost savings should a chatbot provide?How long should it take to make a model understand a company’s data?How many workers can a company displace with AI?OpenAI in 2030 - 26 gigawatts could power between 3.7 million to 17.3 million modern GPU serversOpenAI in 2035 - 50 gigawatts could power between 37 million to 173 million modern GPU serversFEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod