24.03 Liontrust Global Innovation Report - The Rise of AI 04.24 - Flipbook - Page 33
How could this alter the competitive landscape?
What do you think will be the defining criteria of
winners from this AI revolution? What types of
companies run the risk of being disrupted?
I believe that the intersection of technology and biotech
is an inevitability and starting to play out. I am confident that in 10
years or so, the biopharma industry is going to resemble Recursion
more than the other way around. Biopharma is a trillion-dollar industry
where success rates from preclinical stages to the market are less than
10%. There are lots of opportunities for driving greater efficiencies.
I believe that the technology companies have started to move into
this space, leaning in with their strengths. For example, you have
Microsoft with OpenAI – it launched BioGPT. Amazon was involved
with a Phase 1 clinical trial and has acquired companies like One
Medical and PillPack – it is leaning into its strength in logistics and
getting to know the consumer. Oracle acquired Cerner, which I
believe highlights a desire to be a system of record for patient data.
And then you have Google which has incubated companies like
Calico, Verily, and DeepMind’s Isomorphic Labs which was tied to
breakthroughs like AlphaFold.
As biopharma and technology companies evolve, I believe this
biotech space will be dependent upon companies that can build and
command proprietary data sets built for the purpose of training AI/
ML models to uncover novel insights. And that is what Recursion has
been able to build itself – arguably one of the largest such datasets.
What about regulation – how do you view the stance
of regulators towards AI in the healthcare industry?
I think it is incredible to see regulators such as the FDA
being so forward-thinking. When you consider the
potential of LLMs, for instance, I believe that they can immediately
begin assisting regulatory bodies. Imagine LLMs providing a
preliminary review of investigational new drug (IND) submissions.
These tools could provide a first pass, helping to scale the review
process for the burgeoning number of therapeutic potentials. This
kind of technology could be beneficial for regulators in navigating
the complex interplay of science, regulation and commerce.
What are you most excited about in the year ahead?
I believe that this is going to be an impactful year.
Seeing LOWE (our LLM-Orchestrated Workflow
Engine) connect many of our experimental and
computational modules with proprietary and public
data as well as perform next-generation digital chemistry
calculations live, in a non-code way, has been fascinating.
Moreover, I believe that tools like LOWE can help scale the
productivity of teams working in drug discovery where an ability
to code is not a barrier for adoption. I increasingly believe that
the programming language of the future is human, where we will
just talk to a computer as we would with a person to carry out
increasingly complex tasks.
The growing use of AI/ML in drug discovery and development is
not a hypothetical – it is real and right before you – that is what
LOWE represents to me. And I think that sentiment is increasingly
being felt by large pharma companies. You see some companies
like Roche and GSK starting to build their own LLMs. I believe
that this is a notable shift. Complementing Recursion’s software
tools is the integration and scaling of new capabilities within our
platform (the Recursion OS), progression of our partnerships, and
advancement of our pipeline programs through clinical trials and
to data readouts.
The rise of AI: Technology and Innovation Report - 33
