Was just onboarded into Microsoft’s 365 Copilot Frontier Program. I’ll be providing them strategic feedback on potent new AI productivity features they’ll be adding soon. My take: I see CIOs struggle to determine the best path for worker AI uptake. Watch my experiments + learn!

Of all the fast-growing tech products in recent years. @cursor_ai is the fastest to $1B in ARR. This speaks volumes to the demand for AI to help teams + orgs move faster with less cost.

GM chief AI officer steps down www.ciodive.com/news/general... Turovsky is heading off to work on “some exciting new ideas.” My take: Trad firms will have real challenges keeping top talent in an era with sky-is-the-limit AI thinking.

Guardrails and governance: A blueprint for CIOs on responsible generative and agentic AI | @CIO.com www.cio.com/article/4094... My take: Paper policies are ineffective + out. Physical controls and guardrails relentlessly overseen by code are the only real path.

New Research: Are CIOs Ready to Bet Their Agentic Operating Model on Salesforce? futurumgroup.com/insights/are... $SFDC has become an enterprise platform vendor by virtually all measures + is in fact now the #2 software firm, and now #1 in AI agents. My take on whether CIOs will bite.

What is fast becoming the most interesting industry aspect of the Gemini 3 Pro announcement was that it was apparently trained entirely on TPUs. Not GPUs. Not NVIDIA chips. Since by many benchmarks, Gemini 3 Pro is now the leading AI model, this opens up big forks in the road.

1/ I just had an opportunity to dig into Oracle Cloud’s new Multicloud Universal Credits. It’s a fascinating move to become a cloud meta-provider. Oracle is essentially saying that your primary cloud matters less today. Your data + DBs can follow your strategy, not your vendors.

Just took a call from a reporter about new agentic AI announcements from a top tech company. It’s becoming clear that the big tech firms will be able to use their leading/monopoly positions in identity, security, business data, etc. to make their agents the most capable.

Catch my recent appearance on the AWS for Software Companies Podcast | Ep155: Balancing Innovation and Regulation: art19.com/shows/aws-fo... I had a great + wide-ranging convo with @AWSCloud's Massimo Ghislandi on navigating AI adoption and cloud transformation challenges

3 in 10 firms plan to replace workers with AI next year www.ciodive.com/news/AI-layo... Biggest hits expected in IT, financial services and accounting as 1 in 5 CFOs say they've reduce headcount due to AI. The next level of automation is here, and reshaping labor markets.

Overheard: "Sometimes I want to use tokens and sometimes I want to use code [to solve my problem.]" On the growing desire to mix-and-match/blend generative AI with low/pro-code. Especially critical for agents, where you want some determinism. Risk: Killing the 'magic' of AI.

Gemini is sporting a solid and fast-growing number two position vs. OpenAI web traffic shows that the immense investment Google is making paying off. While this is mostly consumer traffic, it shows that significant AI market shifts are still quite possible. Even likely.

#CIOChat Good chat today. Next Chat Topic: Slack/Chat Overload It's been 7 months of our movie tie-ins. Hope you enjoy them. Please suggest topics and invite your IT colleagues. 2pm ET every Thursday!

A3) All of those have their best use cases. #EventDriven architectures can also surface data flow and #API requirements for systems that need to share. Centers of excellence and supply chain of data teams can help when empowered properly. #ciochat

A4) Doubling down on all the related #architecture areas to make sure we get it as right as possible. Strangely, the #sustainability part means we need to keep the coldest data on near-zero-energy storage, which might still be tape in 2026. #ciochat

#CIOChat Q4: The IT endgame: A world where data has mass, ready motion, and intent. CIOs must choose to orbit around hyperscalers or engineer new centers of data gravity. What does a fully sustainable, sovereign, and intelligent data architecture look like in your future state?

Yes, I've just begun to deep dive on photonics. The potential for a 10x gain in processing speed and 30% or more improvements on signal speed through the compute medium is potentially a big deal. Another way we might fight data gravity indeed! #CIOChat

I'm keeping my eyes on a lot of the hybrid electro-optical and silicon photonics folks. Parallel paths, fewer hops, fewer changes of physical layers, etc. can make a big difference. #ciochat

#CIOChat Q3: Modernization often worsens gravity: Every new lake, warehouse, or AI model deepens the pull. What’s your playbook for fighting back? Is it data fabrics, streaming layers, data contracts, metadata gravity assists, or do you simply create virtual data centers?

No question it's pushing the envelope, but data gravity is also pushing around the IT dept, as it can sharply limit what's possible. All this data is terrific, but it's an increasingly heavy weight. The solutions remain limited, unless there's a networking breakthrough. #CIOChat

A1) 💯, not just #IT and #OT creating business data but the explosion of observability data and new forms w/ #AI chat logs, etc. Pushing the #DataArchitecture conversations harder but also tweaking #DataRetention policies end-to-end to keep less but higher value data. #ciochat

#CIOChat Q2: Data sprawl once meant freedom. Now it means friction. Each workload pulled closer to its data reduces multicloud choice and portability. How are CIOs keeping optionality alive when moving data is slower, costlier, and politically/vendor-wise harder than ever?

#CIOChat Q1: As enterprise data volumes explode into petabytes, the sheer scope of IT is changing. Data gravity now determines where apps, AI, and even vendors must live. How is your architecture adapting as data location, not strategy, increasingly dictates what you can do?

I'm joining #CIOChat from Portugal this week, where I'm off the road for a couple of weeks. Looking forward to #reInvent in Las Vegas at the end of the month!

#CIOChat: Welcome to today’s chat. Topic: Fighting Data Gravity

#CIOChat at the top of the hour cc: @cpetersen-cs.bsky.social @jimcrussell.bsky.social @efeatherston.bsky.social @iangertler.bsky.social @jbrodsky.bsky.social @toddmazza.bsky.social @andrewnebus.bsky.social @bhaines0.bsky.social @sappley.bsky.social @amypearlman.bsky.social @jimpeluso.bsky.social

Shadow AI is widespread, and executives use it the most www.ciodive.com/news/shadow-... My take: With more than 80% of workers using unapproved AI tools, a new + better approach to adoption of generative tools must be found.

Join this week’s #CIOChat 2-3 p.m. EST tomorrow. Topic: Fighting Data Gravity

CIOs: Collect the right data now for future AI-enabled services | @cio.com www.cio.com/article/4077... Unfortunately, most orgs should have been doing this for ten years. But there's no time to start like the present.

AI agents worsen IT’s capacity crunch www.ciodive.com/news/AI-agen... Largely as expected, agents that interact with IT systems and call other agents will consume more resources. And will require more sophisticated IT staff to support + debug.

I do like the Hackathon idea. While they are a bit out of favor, I find that done well, they can create real inspiration + motivation for a whole new crop of recent hires. And some existing folks too. #CIOChat

#CIOChat Good chat today. Next Chat Topic: Fighting Data Gravity It’s been 7 months of our movie tie-ins. Hope you enjoy them. Please suggest topics and invite your IT colleagues to the chat. 2pm ET every Thursday

A4) Adopt the #Agile "iterate and improve" mantra for all processes as well as tech. Keep people in the loop on how we're trying to reduce the strain (and waste) everywhere. Throw a #Kaizen-style exercise (or hackathon) in when we can for a burst of hope. #ciochat

#CIOChat Q4: Sustained velocity requires resilience, not adrenaline. What have you learned about resetting pace, rotating workloads, or redesigning org structures so the team can endure multi-year transformation without burning out or breaking culture?

My take: Interviews I’ve had with CIOs and AI practitioners, plus my own experience, shows that AI tooling is a genuine force multiplier. Caution is warranted, especially around creating large tech debt, but considerable (albeit maybe short term) velocity is there. #CIOChat

Not a big believer in the current #AI hype train. Too many folks are tied to both trolley (problem) tracks, but I'm a huge believer in all the fundamentals that make those AI tools a possibility in the first place. #ciochat

#CIOChat Q3: In high-velocity orgs, leadership posture matters more than methodology. How do you typically model calm, clarity, and prioritization amid chaos, so your teams don’t just go faster, but go smarter? Share what rituals, norms, or boundaries actually work.

Yes, the old scalpel to remove everything but the most necessary. It’s a good approach for agile sprint payloads as well. But more delivery capacity is just as key. We now must do much more than before, with ~same resources. This is where AI tools/automation comes in. #CIOChat

A2) Partner up with stakeholders and see if there are any constraints I can relax. It might be as small as buying the team a lunch or a beer and saying thanks more often, but I'd hope that an empowered #CIO with a seat at the strategy table could swing a bit more. #ciochat

A lot depends on the organization and their market position, but they may not be wrong at all. You can only burn the candle with a blowtorch so long until it's time to leave and take people with you if you can. #ciochat

It’s interesting. I’ve had talks w/ several CIOs recently who now think the danger of not acting quickly / responding to change fast is at an all-time high. With AI, disruption is afoot like wild bulls in the street. Certain things can’t seem to wait now. What to do? #CIOChat

A1) Two biggest (sets of) variables for me are desired outcomes and resource levels. Speed for the sake of speed is no good. Racing the time-to-market or time-to-value clock is understandable. And, there's only so much we can do w/o adding resources of various kinds. #ciochat

A1.2) Time-to-value has its own BS to not-BS spectrum of values. Having *something* to appease an unhappy board or shareholders falls in the middle. Can't succeed if we can't keep our jobs... #ciochat

#CIOChat Q2: When velocity is constant, teams have no slack + lose the time to “sharpen the saw”: No retros, refactoring, skill-building, even PTO. What leadership moves help you protect your team’s capacity and humanity when there is no obvious slack in the system?

I’m joining #CIOChat today from Concourse C in Dulles Airport. The sun is setting on a gorgeous day.

#CIOChat Q1: My CIO surveys show ~94% of CIOs are under “high” or “very high” pressure to move faster. Agile, DevOps, low/no-code, vibe coding all promise acceleration, but their velocity can turn into strain. How to decide when speed is strategic vs. when it becomes too much?

I’m joining #CIOChat from Concourse C in Dulles Airport. The sun is setting on a gorgeous day.

A friendly reminder I am also simulcasting #CIOChat on Bluesky, for those of you who prefer to participate there.

#CIOChat: Welcome to today’s chat. Topic: Leadership Under Velocity

Join this week’s #CIOChat 2-3 p.m. EST tomorrow. Topic: Leadership Under Velocity