AI at Scale: A Conversation with Blackstone’s CTO and Global Head of the Operating Team
1. Where does AI create the most value — and where do you see disruptions?
Zemmel: We believe some of the most compelling opportunities today sit close to the infrastructure layer — what Blackstone refers to as the “picks and shovels” of AI — data centers, compute, and related areas tightly coupled to that stack. At the other end, some skilled tasks that current models can already replicate — like parts of content creation or legal workflows — face real pressure.
The biggest opportunity is the vast middle of the economy. Physical businesses, manufacturing, logistics, and retail. These are not “AI-native” sectors, but they are increasingly AI-enabled. From factory layout and robotics to marketing and supply chains, AI is enhancing how real-world businesses operate.
Stecher: As a software engineer, I see the biggest proven impact right now in software engineering’s toolkit. My teams are meaningfully more productive than they were a year ago. And I think that same dynamic will extend into drug discovery, materials science, anywhere the work mirrors software development at its core — design something, test it, refine it, repeat.
2. Software has been one of the most talked-about areas of AI disruption. How is it actually playing out?
Zemmel: Software is too big a category to paint with a broad brush. A simple analytical app layer built on commodity data is going to be easy to replace. A deeply embedded system of record with proprietary data and real workflow understanding is very hard to replace — and possibly has more AI tailwinds than headwinds. Within software, we expect much wider dispersion of returns between companies that get it right and those that don’t.
Stecher: When we diligence software companies, we look at how much proprietary data they have that can’t be replicated from what’s publicly available. That is a moat. If the business is predicated on proprietary people and proprietary data deeply embedded in workflows, AI makes their knowledge base more powerful, not more vulnerable. What’s also shifting is the interface itself — less logging into a website and entering data, more digital assistants that access information directly. Software that that simply adds an interface on top of data that anyone can access is easy to replace. Software that’s embedded in the workflow is not. The other thing I’d watch is the trust factor. Companies that customers feel they genuinely can’t run their business without — those are enduring businesses. Companies that customers tolerate only out of inertia — I don’t think they’re long for this world. New competitors that build on customer trust are going to overrun them.
3. How is AI changing the way Blackstone makes investment decisions?
Stecher: AI gives our people leverage. Think about doing a deal: you’re looking at thousands of documents, and what used to be a weekend-long activity can now be carried out in minutes. That frees our teams to focus on judgment: understanding risk, improving structures, and driving better outcomes for our LPs. [ 4 ]
We’re also using AI at the very top of the funnel. Because of Blackstone’s scale, we see a lot of deal flow. AI allows us to evaluate opportunities faster, say no faster, and spend more time on what we believe are the highest-quality investments. Our private equity and real estate teams are deploying very targeted tools — for example, to help move quicker from raw deal documents to financial models. The goal is simple: eliminate rote work and amplify human insight.
4. Blackstone owns over 270 portfolio companies. How are you driving AI adoption across Blackstone’s portfolio?
Zemmel: We think about our portfolio in three buckets: companies where AI is central, companies where it presents meaningful upside, and companies where it’s relatively immaterial — though that last group is shrinking fast.
We’ve built a dedicated team and a playbook focused on six areas of impact: software engineering, customer experience, combining data to create new insights, content creation, continuous improvement in supply chains, and corporate automation using agents — AI that doesn’t just answer questions but drives the completion of the work end to end. Most of these are already delivering results.
The challenge isn’t necessarily technology — it’s execution. Too many companies run performative pilots that never show up in EBITDA. Our principles are straightforward: CEO-level ownership, clear ROI, data as a competitive advantage, and designing for scale from day one. Let AI help with the routine work so people can focus on what they do best.
5. What about jobs and productivity?
Zemmel: The data is still noisy, but historically technology has created more jobs than it’s destroyed. [ 5 ] The real story is productivity. What software engineers are experiencing today is likely what much of the workforce will experience over the next decade.
Stecher: Lowering the cost of creation expands demand. We saw it with the internet , and we’re seeing it again with AI. There will be real disruption but long term we believe the job impact remains to be seen. And the productivity gains extend well beyond white-collar work, and into really any field where significant time is lost to administrative tasks.
6. AI requires massive energy and infrastructure. How does that translate into opportunity?
Zemmel: Power is the bottleneck. Models are far more efficient than they were early on, down roughly 100-fold since they first launched, but that efficiency has unlocked even more demand. We’re seeing a need not just for energy generation, but for grid modernization and connectivity. Data centers today are often built with years of demand pre-sold — this is about catching up, not building ahead.
Stecher: The data and compute explosion underpins everything. More data trains better models, which drives more usage and more infrastructure investment. Five years from now, AI will be always on — and we believe the infrastructure supporting it will be one of the defining investment themes of the decade.
Accelerating US demand for data centers [ 6 ]
US New Leasing (MW)
7. What makes an AI investment durable rather than speculative?
Zemmel: The dot-com bust was defined by massive fixed costs built ahead of demand. [ 7 ] Today, demand is real and, if anything, underestimated. That doesn’t mean there won’t be mispricing or excess — there will be — but the fundamentals are very different.
Stecher: The cost of building has collapsed. Going from idea to code to working prototype to a customer’s hand is a fraction of what it used to cost — which means more experimentation and more winners, probably from smaller companies. [ 8 ] And when something doesn’t work, the failure is more contained. You’re not writing off a billion-dollar infrastructure buildout — you’re shutting down a project that cost a fraction of that.
Zemmel: Within every sector, the dispersion between companies that execute well on AI and those that don’t is widening. The premium on execution has never been higher. That’s where having an active, hands-on approach to working with management teams really matters.
Capex as a % of revenue [ 9 ]
Important Disclosure Information
This commentary does not constitute an offer to sell any securities or the solicitation of an offer to purchase any securities. This commentary discusses broad market, industry or sector trends, or other general economic, market or political conditions and has not been provided in a fiduciary capacity under ERISA and should not be construed as research, investment advice, or any investment recommendation. Past performance does not predict future returns.
The views expressed in this commentary are the personal views of the authors and do not necessarily reflect the views of Blackstone. The views expressed reflect the current views of the authors as of the date hereof, and neither the authors nor Blackstone undertake any responsibility to advise you of any changes in the views expressed herein.
Blackstone and others associated with it may have positions in and effect transactions in securities of companies mentioned or indirectly referenced in this commentary and may also perform or seek to perform services for those companies. Blackstone and others associated with it may also offer strategies to third parties for compensation within those asset classes mentioned or described in this commentary.