A New Equation for Intelligence Informative Speech
Good morning, everyone. What does it mean to be intelligent? Is intelligence the ability to solve a
difficult equation, recognize a pattern, learn from experience, or make a good decision when the
future is uncertain? We usually think of intelligence as a property of a person or a machine. But
physicist and computer scientist Alex Wissner-Gross proposed a very different way to think about it:
perhaps intelligence can be understood as a force that helps an intelligent system preserve and
expand its future possibilities.
“A New Equation for Intelligence” is a real TEDxBeaconStreet talk by Alex Wissner-Gross, delivered
in November 2013. In the talk, he presents the equation F = T
∇
S
τ
and explains his idea of
intelligence through the concept of causal entropic forces. His central proposal is that intelligent
behavior can be understood as behavior that tends to maximize future freedom of action. This is an
intriguing scientific hypothesis, not a universally accepted definition of intelligence, but it provides a
fascinating framework for thinking about human behavior and artificial intelligence.
The first question is why anyone would want an equation for intelligence. Intelligence is usually
studied through psychology, neuroscience, computer science, and philosophy. Researchers
measure abilities such as reasoning, memory, learning, problem-solving, and adaptation. These
approaches are useful, but Wissner-Gross asks whether intelligence might also be described in
physical terms. Instead of starting with a particular organism or computer program, he asks what
kind of behavior naturally emerges when a system tries to preserve a wide range of possible futures.
The key phrase is “freedom of action.” Imagine that you are walking through a city and suddenly
discover that the road ahead is blocked. Your options become limited. You can turn around, take
another route, stop, or wait. Now imagine that you know several routes in advance and have enough
resources to choose among them. You have more possible futures available. In Wissner-Gross's
framework, an intelligent system tends to act in ways that keep useful options open rather than
becoming trapped in a narrow future.
The equation itself is written as F = T
∇
S
τ
. The symbols come from physics. F represents a force, T
is associated with a temperature-like parameter that determines the strength of the effect, S
represents the diversity of possible future states, and
τ
represents a future time horizon. The
equation describes a proposed relationship in which a system experiences a force toward states that
provide greater future possibilities. It is not a simple formula for calculating someone's IQ. It is a
mathematical expression of a broader theoretical idea.
This idea is connected to entropy, a concept from physics that is often misunderstood. In many
contexts, entropy measures the number of possible microscopic configurations associated with a
macroscopic state. Wissner-Gross's work focuses on a causal form of entropy that considers
possible future trajectories of a system. The important point for this speech is not to treat entropy as
simply “disorder.” Instead, the framework asks how the number and diversity of accessible future
states can influence the behavior of a system.
Why might this produce behavior that looks intelligent? Consider a simple physical system with
many possible paths. If a particular action leaves the system with very few possible futures, it
becomes constrained. Another action might preserve many alternatives. Under the proposed
principle, the second action can be favored because it maintains greater future flexibility.
Surprisingly complex behaviors can emerge from this kind of simple underlying rule, including
behaviors that resemble planning or exploration.
Wissner-Gross has used computer simulations to explore this idea. In his talk, he presents examples
involving physical systems and shows how behaviors that appear surprisingly purposeful can
emerge without explicitly programming those systems to pursue particular human goals. The
significance of these demonstrations is not that a simulation has suddenly become conscious.
Instead, they illustrate how goal-like behavior might emerge from a general physical principle rather
than from a long list of manually programmed instructions.
One example discussed in this context is tool use. An intelligent system that can manipulate objects
may gain access to more possible future states than a system that cannot. A tool can extend
physical capabilities and create new options. Social cooperation can have a similar effect. Working
with other agents can increase the range of actions that are possible for an individual or group.
These behaviors can therefore be interpreted as strategies for expanding future possibilities.
This perspective has interesting implications for artificial intelligence. Traditional computer programs
are often designed to achieve specific objectives. An AI system might be instructed to play chess,
recognize images, translate text, or recommend products. But general intelligence is much broader.
A general intelligent system needs to handle situations that its designers did not specify in advance.
A principle centered on maintaining future options offers one possible way of thinking about how
adaptive behavior could emerge across many different tasks.
However, we should be careful about what the equation actually proves. Intelligence is a complex
concept, and there is no scientific consensus that one equation completely explains human
intelligence. Human cognition involves brains, bodies, learning, memory, perception, language,
emotion, social interaction, culture, and development. A physical theory can provide an interesting
perspective without replacing the many other scientific approaches used to study intelligence.
There is also an important distinction between intelligence and consciousness. A system can
behave adaptively without necessarily having subjective experiences. A thermostat responds to
temperature, but we do not normally consider it conscious. A sophisticated artificial system can
make complex decisions without proving that it has feelings or an inner point of view.
Wissner-Gross's framework is primarily about behavior and physical dynamics, not a complete
theory of consciousness.
The idea of maximizing future freedom also connects to everyday human decision-making. When we
learn a new skill, save money, maintain our health, build relationships, or acquire knowledge, we
often increase the number of choices available to us later. Education, for example, can open career
and intellectual possibilities that would otherwise be unavailable. In this sense, intelligent behavior
can sometimes be understood as investing in future flexibility rather than optimizing only for the
immediate moment.
Perhaps that is the most useful lesson from this unusual equation. Intelligence may not always look
like knowing the answer immediately. Sometimes intelligence means creating better choices for
tomorrow. It means exploring, learning, adapting, using tools, cooperating with others, and avoiding
situations that unnecessarily eliminate future possibilities. Whether or not Wissner-Gross's equation
becomes a complete theory of intelligence, it offers a powerful question: What if intelligent behavior
is fundamentally about keeping the future open?
A new equation for intelligence does not give us a final answer to what intelligence is. Instead, it
gives us a new way to ask the question. By connecting intelligence with physical systems, entropy,
adaptation, and future possibilities, Alex Wissner-Gross challenges us to look beyond traditional
definitions. Intelligence may be measured in many ways, but one remarkable feature of intelligent
behavior is the ability to avoid being trapped by the present and to create more possibilities for the
future. Thank you.
