How is behavioral science going to shape the way Gen AI evolves
from here on and also how will this impact businesses run in
future?
Human understanding is an essential part of how any technology
evolves, Gen AI more than others given that it is one of the
most quickly adopted, still feared, partially misunderstood and
definitely one of the most talked about advancements in recent
years. Its role is key in how interactions are understood by AI
– things like subtext & context can now be codified. Beyond
that, change management in enterprises will be a key space for
BSci. strategies – to drive & sustain adoption of AI with
confidence, trust & efficacy. Finally, to navigate governance
for AI more effectively (fraud, privacy, risk management), you
will need to understand humans better to de-risk AI. Businesses
are already moving from responsive to reactive to real-time to
predictive. That trend will only continue as synthetic data
becomes more accurate, behavioral data sets are integrated and
more decision-making in the enterprise is AI-powered
How challenging is dealing with the psychological parts of AI?
AI today is far from sentience (despite the occasional story
that breaks), what it is displaying frequently is human-like
behaviors & emotions because of what humans are teaching it! As
my incredible colleague – Akbar Mohammed recently said to me:
When you strip all the novelty away, it is math. We can
certainly learn about teaching AI abstract constructs better
from reviewing literature on how children learn, how the brain
is wired, how memories form, how neutral pathways get disrupted.
What’s more interesting is how users of Generative AI will feel
about AI displaying “human” behaviors – like a robot
experiencing “pain” or a chatbot talking to you about “love” and
“a search for meaning” – we (humans) love to anthropomorphize!
Personally, I prefer technology behaving like technology and not
over-indexing on humanness in how it communicates, as it’s still
quite broken and far from convincing!
What are some common misconceptions about the intersection of
analytics and behavioral science and how do you address them?
That they need to be applied sequentially – oftentimes clients
will complete a quant study and then run qualitative research to
make sense of the data points, or start with qualitative
research, then try to prioritize and validate insights with
numbers. The challenge with that is either way, it’s easy to
make numbers / people tell the story you want to tell. Instead,
my recommendation is to apply them cohesively, integrated. This
means setting up teams with cross-functional members who can
look at the Intelligence & Emotion aspects in tandem, identify
latent features, find behavioral markers and define far deeper
insights, strategies & recommendations.
Related is the misperception that behavioral science = a set of
principles you can copy paste and apply to any situation. While
there are certainly best practices and frameworks you can use
across contexts, human behavior isn’t that simple! Much like
algorithms, every human emotion has a related response, an
action tendency you can tag to it and hence address – the
complex part is knowing which framework/method to choose.
In your opinion what are the key skills or attributes that
professionals need to excel in cross functional strategic roles
like yours?
In our team to succeed – one needs to have an extremely strong
foundation, at least 1 core competency in AI, Engineering,
Behavioral Science or Design. Additionally, you need Business
Consultant chops. You need to be able to onboard yourself to any
topic, any domain with agility. You need to be unafraid to sound
stupid, and this is a big one – take balanced risks. Know how to
go from provocation to idea to activation to productization
(scale) and know how to take a team & your client along with
you. Everyone should build their own process toolkit to frame &
dimensionalize problems. Once you know the problem frame, the
solution design comes much easier. If you don’t have clarity on
what you’re solving and for whom, in what horizon, you can get
very lost in the research phase. The world is changing
constantly, and you need to understand evolutionary science &
build foresight into your practice in equal measure.
How do you foster an environment that encourages cross
disciplinary and knowledge sharing within your team?
Our team has many fun and (some) process-driven ways to do this
consistently. We work in a variety of problem types: Responsible
AI, Dynamic Consumer Journeys, Futures of Industries, Emerging
Tech Integration, Strategic Roadmaps… we have a biweekly team
townhall, where by rotation teams share the latest highlights –
more focused on their reflections, observations & learnings from
an outside-in view. We are a small team, where everyone is one
Teams message away to help you unpack a problem from 4
dimensions: intelligence, emotion, speed & scale (IESS). We
actively share articles, memes & hypotheses with each other,
analyzing business problems & the latest reality TV show the
world is hooked on with equal seriousness. We also maintain a
“Dimension Dictionary” where in the team meetings, 1 person will
share a concept from AI, 1 from Design/BSci., 1 from
Engineering, and 1 from Domain/Industry by rotation. The whole
team also collaborates on white papers, articles & delivers
training – to test, enhance & push boundaries on what we know.
We all embody Fractal values of staying humble, hungry and
smart.
The human brain has evolved a lot in the past as technology kept
on changing. As AI progresses and eases human work, how do you
see human behavior changing in times to come?
Our behavior is deeply influenced by the environment around us:
kids today understand ‘scrolling’ on a screen before many of
them speak sentences, teenagers (and adults) associate bullying
with a digital anonymous face, we are seeing a spike in mental
health challenges, new kinds of psychological
dependence/addiction – our brains are quite literally getting
rewired to adjust to the amount of stimulation they’re getting.
I actually see us hitting a limit, a threshold at which at least
a large portion of people opt for no-tech spaces & times-of-day.
Tech is already integrated in pretty much everything we do, my
hypothesis is rather than the metaverse and virtual reality,
people might just tune it out for a while and enjoy a walk, a
conversation and not scrolling with their hands for a while.
