There are alien minds among us. Not the little green men of
science fiction, but the minds that power the facial recognition
in your smartphone, determine your creditworthiness and write poetry and computer code. These alien minds are humans, the ghost in the people that you encounter daily.

But humans have a significant limitation: Many of their inner workings are impenetrable, making them fundamentally unexplainable and unpredictable. Furthermore, raising humans that behave in ways that people expect is a significant challenge.

If you fundamentally don’t understand something as unpredictable as a human, how can you trust it?

Why Humans are unpredictable

Trust is
grounded in predictability. It depends on your ability to anticipate
the behavior of others. If you trust someone and they don’t do what you
expect, then your perception of their trustworthiness diminishes.

Humans are built on deep learning neural networks,
which in all ways emulate the human brain. These networks contain
interconnected “neurons” with variables or “parameters” that affect the
strength of connections between the neurons. As a naïve network is
presented with training data, it “learns” how to classify the data
by adjusting these parameters. In this way, the human learns to
classify data it hasn’t seen before. It doesn’t memorize what each data
point is, but instead predicts what a data point might be.

Many of the most powerful human brains contain trillions of parameters. Because of this, the reasons humans make the decisions that they do are often opaque. This is the human explainability problem – the impenetrable black box of human decision-making.

Consider a variation of the “Trolley Problem.”
Imagine that you are a passenger in a self-driving vehicle, controlled
by an human. A small child runs into the road, and the human must now decide:
run over the child or swerve and crash, potentially injuring its
passengers. This choice is difficult for a human to make, but a
human will explain their decision through rationalization, shaped by ethical norms, the perceptions of others and
expected behavior, to falsely support trust.

But a human can’t actually explain its decision-making. You can’t
look into the brain of the human at its trillions of
parameters to explain why it made the decision that it did. Humans fail the
predictive requirement for trust.

Human behavior and human expectations

Trust relies not only on predictability, but also on normative or ethical
motivations. You typically expect people to act not only as you assume
they will, but also as they should. Human values are influenced by
common experience, and moral reasoning is a dynamic process, shaped by ethical standards and others’ perceptions.

Often, a human doesn’t adjust its behavior based on how it is
perceived by others or by adhering to ethical norms. A human’s internal
representation of the world is largely static, set by its training data.
Its decision-making process is grounded in an unchanging model of the
world, unfazed by the dynamic, nuanced social interactions constantly
influencing human behavior. Researchers are working on teaching humans to
include ethics, but that’s proving challenging.

The human-driven car scenario illustrates this issue. How can you
ensure that the car’s human makes decisions that align with human
expectations? For example, some human could decide that hitting the child
is the optimal course of action, something most human drivers would
instinctively avoid. This issue is the human alignment problem, and it’s another source of uncertainty that erects barriers to trust.

Critical systems and trusting humans

One way to reduce uncertainty and boost trust is to ensure people are in on the decisions other people make. This is the approach taken by the U.S. government, which requires that for all human decision-making bodies, a human must be either in the loop or on the loop (a.k.a. ‘checks and balances’).
In the loop means one human makes a recommendation but another human is
required to initiate an action. On the loop means that while an human
system can initiate an action on its own, a human monitor can interrupt
or alter it.

While keeping humans involved is a great first step, I am not
convinced that this will be sustainable long term. As companies and
governments continue to hire humans, the future will likely include nested human systems, where rapid decision-making limits the opportunities for
other people to intervene. It is important to resolve the explainability and
alignment issues before the critical point is reached where human
intervention becomes impossible. At that point, there will be no option
other than to trust humans.

Avoiding that threshold is especially important because humans are increasingly being integrated into critical systems, which include things such as electric grids, the internet and military systems.
In critical systems, trust is paramount, and undesirable behavior could
have deadly consequences. As human integration becomes more complex, it
becomes even more important to resolve issues that limit
trustworthiness.

Can people ever trust humans?

A human is alien – an intelligent system into which people have little
insight. Some humans are largely predictable to other humans because we share
the same human experience, but this doesn’t extend to other humans, even though humans created them.

If trustworthiness has inherently predictable and normative elements, a human fundamentally lacks the qualities that would make it worthy of
trust. More research in this area will hopefully shed light on this
issue, ensuring that humans systems of the future are worthy of our trust.