WEBVTT

00:00:00.625 --> 00:00:03.586
Tobias, what is the role of applied AI?

00:00:03.586 --> 00:00:06.840
So the core mission of
applied AI at Anthropic

00:00:06.840 --> 00:00:11.594
is to translate this rapidly
increasing model capability

00:00:11.594 --> 00:00:15.640
into actually real tangible
value for businesses.

00:00:15.640 --> 00:00:16.516
And why that's important

00:00:16.516 --> 00:00:18.476
is because the model capability,

00:00:18.476 --> 00:00:22.063
we're seeing it increase
at this exponential rate.

00:00:22.063 --> 00:00:26.317
This is task complexity horizon
that these agents can run,

00:00:26.317 --> 00:00:29.696
but the business value
realized is lagging behind.

00:00:29.696 --> 00:00:32.407
And that is the gap
that the applied AI team

00:00:32.407 --> 00:00:33.616
is looking to close.

00:00:33.616 --> 00:00:36.369
And how are you working with
partners to close that gap

00:00:36.369 --> 00:00:37.620
and deliver value?

00:00:37.620 --> 00:00:38.955
So the reality, Tom,

00:00:38.955 --> 00:00:41.916
is that the majority of enterprises

00:00:41.916 --> 00:00:46.212
will experience AI adoption
through our partner ecosystem.

00:00:46.212 --> 00:00:48.131
And the reason for this is twofold.

00:00:48.131 --> 00:00:52.385
Number one is the scale and
then number two is the depth.

00:00:52.385 --> 00:00:55.930
The scale of the enterprise
demand far outstrips

00:00:55.930 --> 00:00:58.808
the number of technical
resources we can staff up.

00:00:58.808 --> 00:01:00.727
So that is where we are looking

00:01:00.727 --> 00:01:03.271
to bring in the technical teams

00:01:03.271 --> 00:01:06.608
at partners like BCG
to help drive adoption.

00:01:06.608 --> 00:01:09.486
And the industry depth

00:01:09.486 --> 00:01:12.989
and industry knowledge that
your organization brings,

00:01:12.989 --> 00:01:14.949
that these partners bring, helps

00:01:14.949 --> 00:01:19.454
to actually realize AI
adoption in these enterprises.

00:01:19.454 --> 00:01:22.957
A perfect example of this
is with Claude security,

00:01:22.957 --> 00:01:25.960
where we're seeing demand across the board

00:01:25.960 --> 00:01:28.463
for how can organizations shore up

00:01:28.463 --> 00:01:31.549
their software vulnerabilities

00:01:31.549 --> 00:01:36.429
in this ever-increasing
cybercapability development.

00:01:36.429 --> 00:01:37.388
And so that's where we're partnering

00:01:37.388 --> 00:01:39.599
with select firms like BCG

00:01:39.599 --> 00:01:41.976
to actually bring that to enterprises.

00:01:41.976 --> 00:01:43.144
And as you look ahead,

00:01:43.144 --> 00:01:46.022
what are you most excited
about for the future of AI?

00:01:46.022 --> 00:01:48.608
For the industry transformation,

00:01:48.608 --> 00:01:50.610
I would say that health care continues

00:01:50.610 --> 00:01:53.530
to be one of the most exciting outcomes.

00:01:53.530 --> 00:01:55.657
And I'm very excited to see enterprises,

00:01:55.657 --> 00:01:59.285
if it's in drug discovery
or rare disease research,

00:01:59.285 --> 00:02:01.037
continue to push the limits there.

00:02:01.037 --> 00:02:03.331
The technical evolution,

00:02:03.331 --> 00:02:04.833
I'm really excited about

00:02:04.833 --> 00:02:08.878
multi-agent orchestration
becoming the norm

00:02:08.878 --> 00:02:12.924
and also trying to shift
from a human-in-the-loop,

00:02:12.924 --> 00:02:15.176
where you have to hand off every task,

00:02:15.176 --> 00:02:18.012
to actually having these
agents be more proactive

00:02:18.012 --> 00:02:19.722
around kicking off jobs on their own.

00:02:19.722 --> 00:02:22.058
Tobias, thank you very much
for spending time with us.

00:02:22.058 --> 00:02:23.351
Thanks for having me, Tom.