Forum recap | Fireside chat with Nicholas Thompson wants you to know about AI, power, and the future of being human
The Atlantic CEO offered a candid, compelling glimpse into how AI is reshaping power, privacy, and the professions we thought were safe.
A key topic at Forum 2025: AI. While most of Forum focused on AI in specific multifamily use cases—leasing and operations—this closing fireside chat featuring Nicholas Thompson, CEO of The Atlantic, zoomed out to the broader implications and questions about AI. Where is it taking us? Who will control it? What might it do to (or for) humanity along the way?
The under hyped, the overhyped, and the unspoken dangers
Christiansen opened with a direct question: What’s the most under appreciated risk of AI—and what’s the most overhyped?
The under appreciated risk
Our collective grip on reality is about to get fuzzy. “It won’t be long before we have no idea who’s real and who’s not,” Thompson said. Whether on social media or in a chatbot conversation, the line between human and machine is disappearing. “And we don’t know how humanity can function in that world. We don’t know if democracy can function.”
The second under appreciated risk? The quiet, subjective power of the people behind the algorithms. “The folks who control the algorithms set the terms,” Thompson said, recounting a story of a parole AI model that had been—intentionally—created to release people earlier. The message: it’s not just about bias; it’s about intent. The power to subtly shape perception, policy, and public opinion is no longer just in the hands of publishers or governments. It’s in the code.
The overhyped risk
“According to one report, there’s a 70% chance AI will kill us all by 2029.” said Thompson. “I hope that’s overhyped.” Jokes aside, he emphasized that while existential risks make headlines, the more immediate danger is less cinematic and more systemic—slow-creep distortions in trust, truth, and economic power.
Follow the value, follow the power
Beyond the risks, one of the most urgent and unresolved questions in AI today is where value—and power—will ultimately concentrate.
Billions are invested across the stack: in hardware and chips from companies like NVIDIA, in the development of foundation models, and at the application layer where tools like Sierra are transforming how businesses operate. The landscape is crowded, but the stakes are clear: whoever owns the most leverage in this ecosystem could shape not just markets, but influence at scale.
“Right now, the value accrues to the marketers who can raise the most money based on AI,” said Thompson. Valuations are outpacing revenue, and the hype cycle is in full spin.
But longer term? He fears a consolidation at the foundation model layer. “A small number of companies that have foundation models will control most of the value. There’ll be layers and skins built on top, but that core will be hard to dislodge.”
And here’s the kicker: once a model gets ahead, it can design the next version of itself. “A small advantage becomes a big advantage,” Thompson said.
In this framework, power doesn’t just correlate with value—it compounds with time.
Who’s disrupted first?
As AI reshapes industries, which professions are most at risk? Short answer: the ones closest to the training data.
“Coders, journalists… we’re the most screwed,” Thompson said, half-joking, half-not. Professions built on written, structured output are easier for AI to replicate than roles involving creativity, social interaction, or legal/cultural nuance. “That’s your moat,” he said to the audience of multifamily operators. “Property management still has face-to-face negotiations, complex human relationships.”
That said, AI will transform parts of every industry—including leasing. Data analyst roles will shift dramatically. Centralized AI agents will handle more interactions. But the core human component? That remains sticky and aligns with the underlying message woven throughout all of the Forum programming: AI + YOU = the future of multifamily
At The Atlantic, Thompson is proactively widening that moat: “We prioritize new facts, deep reporting, style and grace, and building human relationships. AI can’t replicate those—at least not yet.”
Teams and companies who use AI will outperform those who don’t
In multifamily, as in many industries, the competitive edge isn’t just about having AI—it’s about knowing how to use it. The difference between companies that unlock real efficiency and those that get stuck in experimentation often comes down to one thing: adoption at the team level.
Thompson sees the same dynamic playing out in journalism, where the professionals who lean into AI are pulling ahead of those who don’t.
His staff initially balked when he struck a deal with OpenAI. “They stole our data—why are we making a deal with them?” they asked.
But Thompson’s take was pragmatic: if your data is being used, and one company is offering compensation and collaboration, why not build a bridge?
Today, his position is clear: “AI won’t beat journalists. But journalists who use AI will beat those who don’t.”
To that end, his team has leaned in—inviting OpenAI to train staff, launching Slack channels for experimentation, and embedding AI into ad pitches and editorial planning. Just like leasing teams who in the beginning might have viewed AI, as a competitor, not a collaborator adoption and team sentiment around these new solutions adapts both based on use, and on change management initiatives and feedback loops like those the Windsor team discussed in their presentation on change management.
Privacy, trust, and the need for better tools
For an industry that asks for a lot of personal information to screen applicants—both privacy and fraud aren’t theoretical concerns. It’s a daily operational risk.
Thompson didn’t sugarcoat it regarding privacy: “Don’t trust the companies,” he said, when it comes to how tech platforms handle your data. His advice? Use enterprise-grade tools with strong data governance, adjust your settings, and never upload anything sensitive you wouldn’t want swept into a training set. “The history of the tech industry is that they always end up figuring out a way to get it.”
For multifamily operators navigating AI integrations—from fraud screening to leasing chatbots—that warning lands close to home. The push to modernize workflows must be balanced with airtight security practices, especially when it comes to personally identifiable information (PII) and compliance with an evolving regulations.
In an environment where even images, voices, and messages can be spoofed, Thompson believes trust itself is becoming a business differentiator. “Whoever builds human authentication systems that really work—that’s a great business to be in,” he said. In multifamily, that could be the difference between securing a lease… or falling victim to fraud.
Will AI make us more human—or less?
Imagine a digital world so perfectly tuned to your preferences that real life starts to feel… inconvenient.
Thompson warned that AI could usher in just that: hyper-personalized, simulated realities designed to satisfy every desire.
That future isn’t just unsettling—it’s dangerous.
As AI becomes more engaging, the risk isn’t killer robots; it’s losing touch with real relationships, real community, and real work. As Thompson put it, companies that “do not love you, but do want your money” can now entertain you infinitely—and that’s where humanity starts to slip.
But there’s an alternate path. Citing insights from the Vatican’s AI advisory group (yes, really), Thompson offered a surprisingly hopeful view: AI is a human tool. Not something apart from us, but something that can reflect and even reinforce what makes us human—if we use it with intention.
For multifamily operators, the message is clear: AI won’t replace human connection, but it can make your teams more effective if you preserve what renters value most—authenticity, empathy, and trust.
The future isn’t bots versus people. It’s AI plus humans, delivering experiences no algorithm can fake.
The takeaway for multifamily leaders
So what does all this mean for property management professionals navigating AI in their own business?
It means the tools you choose—and how your teams use them—will shape your future.
It means your moat is human interaction, trust, and transparency.
And it means that while AI might commoditize certain roles or tasks, it also amplifies the capabilities of those who learn to wield it well.
“Companies that are smart about AI will get 5–10% better,” Thompson said. “But the real game-changer will be the one that’s AI-native and wipes out an incumbent.”
That’s a call to arms for industries, and individuals, to not just adopt AI—but to rethink how we lead, build, and connect in its presence.
In a world of machines, our humanity might just be the most valuable differentiator of all.