How to greet the dawn of “future-state predictive intelligence”
- Dataminr’s proprietary LLMs analyze real-time events at such speed that whole new strata of planning and strategy become possible.
- The company offers “real-time visibility into the market — and the world — to make better and more informed decisions.”
- Here, Dataminr president and COO Brian Gumbel discusses risk, “convergence crises,” unintended misinformation, and more.
Cybersecurity veteran Brian Gumbel — president and chief operating officer (COO) at Dataminr — works at the confluence of real-time information and AI. Mainlined into humanity’s daily maelstrom of data, Dataminr detects events “on average 5 hours ahead of the Associated Press” — it picked up the 2024 Baltimore bridge collapse, for example, about an hour ahead of all mainstream media sources. The accuracy rate of its “news” is, says Gumbel, a highly impressive 99.5%.
On top of that, the company’s AI systems, built on proprietary LLMs, analyze real-time events at such unprecedented speed that whole new strata of planning and strategy become possible. Soon, predicts Gumbel, “information will be automatically generated and updated as events unfold, enabling us to anticipate what is likely to happen next and take proactive steps to address it.”
While nobody is claiming anything equivalent to Philip K Dick’s future-crime-perceiving “precogs” in The Minority Report, the technology is nonetheless exciting. “The future is unfolding right now,” Gumbel says. “This isn’t the realm of science fiction.” Even though the focus on analyzing real-time information may appear niche, the relevance here to businesses across the board is clear. Any CEO would benefit from “real-time visibility into the market — and the world — to make better and more informed decisions.”
Here, Gumbel joins Big Think to explore preemptive responses to risk, why we should prepare for a “convergence crisis,” a solution for unintended misinformation in our organizational silos, and more.
Big Think: Dataminr’s company mission is to “transform the future of real-time information.” Five years from now, what are a couple of the most transformative changes we could see?
Gumbel: We are seeing profound changes already and don’t have to wait five years for these to be more broadly realized. The use of real-time information is shifting from its current security use case, where it’s often siloed in the Global Security Operations Center, to a much broader business use case across the entire organization. It is becoming less acceptable for leaders to rely on historical data and trend analysis to make critical business decisions; they want always-on, real-time visibility into the market — and the world — to make better and more informed decisions.
Additionally, the use of real-time information will evolve from current-state situational awareness to future-state predictive intelligence. Using promptless AI, information will be automatically generated and updated as events unfold, enabling us to anticipate what is likely to happen next and take proactive steps to address it. We call this PreGenerative AI (PreGenAI), and it’s possible due to Dataminr’s immense event archive, containing the unique arc of how every historical event transpired over the last 15+ years. Our AI platform is able to train on all this data and then anticipate potential trajectories and effects of complex developments as they evolve in real time.
Big Think: Will our ability to predict events ultimately feel like sci-fi?
Gumbel: Predicting events 10 years from now might still be out of reach, but cutting-edge AI is already here, reshaping how we respond to disruptions. Today, we have the tools to catch the earliest signals of risks and act before threats fully take shape. Just imagine something like a major supply chain disruption — say, a weather event or natural disaster significantly disrupts shipping — and PreGenAI could anticipate the next potential impacts to your business so you can plan and respond proactively. This would be utterly transformative in how organizations approach decision-making amid unexpected events and crises. The ability to continuously model what’s next is transforming decision-making as we know it, and the future is unfolding right now. This isn’t the realm of science fiction; it’s reality, accelerating at breakneck speed.
Big Think: In what other ways has AI impacted event analysis, and how can it alter the speed of our actions in the real world?
Gumbel: There is no doubt that the pace of event detection and mitigation has changed dramatically with AI. With ReGenAI, in particular, events aren’t just detected and described once in an alert; relevant information and context are updated in real time, over and over as the situation unfolds, without even being prompted by the user. That’s a big shift in how teams are able to respond to crisis situations. We’re moving from “what happened” to “what’s happening right now,” which gives those responsible for taking action such a huge advantage. That kind of speed doesn’t just change action, it changes behaviors, giving people the confidence to make faster, more informed decisions. The upside of that is we can preemptively respond to risk in ways never before possible. It’s a new kind of clarity that enables empowered decision-making.
Big Think: How do you define a “convergence crisis,” and why should companies be concerned?
Gumbel: In today’s hyperconnected world, and [with] the volatility of the risk landscape, there’s rarely a digital event that doesn’t have an effect on the physical world, and vice versa. The effects of something like a data breach often don’t stay within the confines of bits and bytes — they can trigger an equally (or more) severe problem in the physical world. This is the “convergence crisis” companies should prepare for.
The ability to continuously model what’s next is transforming decision-making as we know it.
In addition, digital crises resulting from physical events can be catastrophic. Natural disasters like hurricanes or wildfires can cause cyber disruptions by taking out power lines or damaging equipment. A flooded data center can bring down entire businesses. Organizations must understand where these converging risks are and take a proactive approach towards remedying and preventing their impact.
Big Think: What are some of the tactics companies can employ to handle a convergence crisis?
Gumbel: The most impactful change companies can make is undoing the traditional divide between cybersecurity and physical security teams. Instead of each operating under its own silo, companies need to establish a holistic security approach that views risk across all domains, even incorporating operational, reputational, legal, and human resource teams. The goal is to evolve from a reactive posture to one of true business resilience, where holistic security is seen as a unified, board-level imperative in which companies take several proactive measures.
AI is the enabler for corporate security teams to transition from a reactive to a proactive defense. Early attempts at a proactive approach generated oceans of irrelevant data that cluttered the decision-making process. AI, on the other hand, can cut through this noise with real-time discovery of events to deliver a clear, fast, and accurate understanding of newly formed risks and threats.
Big Think: Does the arc of AI and data bend towards the eventual elimination of misinformation?
Gumbel: The sheer volume and velocity of online content, combined with evolving tactics by bad actors, make the full elimination of misinformation unlikely. Plus, the complexity of verifying real-time information adds to the challenge. Misinformation adapts as technology evolves, so it’s a problem we’ll need to continuously battle.
The goal is to evolve from a reactive posture to one of true business resilience.
There is another form of misinformation that is less discussed, but where I believe we can make significant inroads — the unintended misinformation that forms in organizational silos. At Dataminr, we champion transparency by pushing relevant, real-time information across all key business functions to eliminate information asymmetry, create shared understanding, and allow for faster responses. Silos are reduced or even eliminated when all parts of the organization are working with a single source of truth based on accurate, real-time information, empowering all parts of the organization with more informed decision-making.
Big Think: Are there any influential figures in AI whose work — or thought-leadership — you admire?
Gumbel: There are many, but Yann LeCun comes immediately to mind for his pioneering work in AI, especially in deep learning. His research into neural networks has transformed how machines process information, which has been vital to the development of Dataminr’s proprietary LLMs and our ability to harness AI for real-time information discovery. LeCun has done exceptional and influential work to move us all forward.
