19552 GW Mag 26_Final_singles - Flipbook - Page 29
economic data like unemployment rates, GDP growth and
inflation figures. She also had AI reflect the political pressure
President Donald Trump exerted on Powell to slash rates.
The AI committee nearly matched the real-life policymakers’
decisions. Both the AI and actual Fed kept interest rates
unchanged in a range of 4.25 precent to 4.5 percent. The
AI version landed at a 4.42 percent midpoint.
Perhaps more impressively, the AI simulation recorded
rare dissent among Fed members. In reality, two members
defied the rest of the committee and voted for lower rates—
the first dual dissent for rate cuts since 1993. Applying the
political influence ingredient, Sinclair explained, swayed
arguments and fragmented AI personas in the same direction
as their real-life counterparts.
“This simulation shows that the Federal Reserve is only
partially insulated from politics,” Sinclair noted in a working
paper. “Outside scrutiny can shape internal decision-making,
even in an institution guided by formal rules.”
In its current format, Sinclair stresses that the model is a
simulator, not a forecaster. For now the goal is to grasp how
the committee sets policy, not necessarily predict its decisions
in advance. Still, for economists, Fed watchers and market
trackers “it helps us better understand how all of these
mechanisms work, what [the Fed] is thinking and what
kinds of influences are at play,” she said.
Similarly Assistant Professor of Interior Architecture
Douglas Crawford guides students in his classes through a
broad ecosystem of AI platforms, some of which are tailored
to interior design workflows. General-purpose tools like
ChatGPT and Claude, for example, enable his students to
more seamlessly move between concept development and
graphic production, he said. Meanwhile, hybrid representation
techniques can aid in classroom projects like enhancing
photographs of students’ physical models by adding lighting
effects, scale figures and material detail. “AI is no longer a
novelty in the classroom,” Crawford said. “It is an integral
tool that is reshaping how students engage the design
process, from initial concept through final representation.”
Economics’ Tara M. Sinclair created an AI simulation of Federal Reserve
meetings to offer a window into their decision-making processes.
A Seat at the Table
Despite his AI apprehension, art therapy’s Potash saw
his interdisciplinary NIH-funded project as an opportunity
to expand the frontiers of his field. Since its founding in
1971, the CCAS Art Therapy program, one of the first of its
kind in the country, has been consistently recognized as an
innovator. “The idea that we are still pushing our profession
forward is exciting,” Potash said.
As part of the project, Potash taught a summer 2025
course in which 10 GW art therapy students were paired with
clients from the Specially Adapted Resource Centers (SPARC)
in Virginia. The students guided people with profound
disabilities through writing prompts as they generated AI
With AI “we can create [Federal
Reserve] personas with flexible
personalities that give us a close
simulation of the real person.”
— TARA M. SINCLAIR, PROFESSOR
OF ECONOMICS
images in therapy sessions. The class is currently completing
a manuscript detailing what Potash called “one of the few
research examples of AI art making in art therapy.”
The project raised larger questions about the ethics of
inserting AI into therapeutic settings. Students debated
whether AI art is a “shortcut to creativity,” Potash said, noting
that art therapy is as much about the experience of making
art as the results. “The creative process is a way to work
through one’s ideas, express oneself, come up against
challenges, recognize new strengths,” he said. “The
struggle is part of it.”
Still, Potash sees the potential for expanding AI-aided
art therapy among clients like veterans and people with
Parkinson’s disease and cognitive or physical limitations.
And while he still has reservations, he plans to stay engaged
in the discussion.
“I could foresee a situation . . . where we as a field have
to react to [AI] as opposed to helping shape it around those
intentional and ethical considerations,” he said. “I want to
be at the table having those conversations.”
GW Arts & Sciences | RESEARCH & LEARNING | 29
