00:00Let me just first get your take on the Anthropic Pentagon debate because it highlighted the importance of, I guess,
00:08government control versus the safety of artificial intelligence. How do you see it?
00:13For us at Mr. Lea, we are very much focused on open weight and open source technology.
00:19We believe in equipping really the builders in having the control of the technology they need to be able to
00:25transform their business, particularly in the private sector.
00:28We are in Bloomberg, invest in financial services institutions.
00:32So for us, what matters is empowering the builders really to have the technology and the large language model they
00:37need to achieve their business outcomes.
00:40What role, if any, do you think government should have in shaping how AI systems are formed throughout the world?
00:45Yeah, from a regulation perspective for us, it's all about giving to those enterprises the controls they need to really
00:52help enable really the systems that they are trying to build.
00:56So for us, the core of that is really the open source approach to how the technology can really be
01:02spread around the world and really battle test that technology for their use case.
01:07I was talking with an entrepreneur yesterday who said, you know, when I use AI at my company, I want
01:12control of the prompts.
01:13You know, I want my own privacy and to be able to hold the technology on my servers.
01:21How much do you hear about that from the customers that you work with?
01:24It's such a great question.
01:26I think larger enterprises have noticed that the ROI of chatbots in their enterprise is a little bit limited, and
01:34they are really trying to deeply transform their business.
01:36In order to transform their business, the theme of portability, being able to deploy those systems on-premise, on your
01:43own walls, but also the topic of customization, making those systems their own, customizing the model to their data becomes
01:50very important.
01:50So a lot of the discussions we're having with large banks, such as HSBC or BNP Paribas, are all about
01:57gaining that control so that they can customize the models really to what they are trying to achieve within their
02:02walls versus just using a chatbot to empower employees.
02:05By the way, Marjorie, I'd assume that one of the most common questions you get are, are your customers saying,
02:10thanks for this AI, we're now going to go fire a bunch of people.
02:13I'm sure you get that a lot.
02:14Has there been any conversation about the enterprises you're working with, about their ability to either reduce or maybe not
02:20grow headcount as much with the use of your tools?
02:22To us and where we are today from a maturation perspective of technology, there's two ways to measure return on
02:29investment for those businesses.
02:30One is actually creating new revenue opportunities.
02:33That technology enabled those companies to do things they were not able to do before, so creating new revenue streams
02:39is the first topic.
02:41The other topic is definitely improving efficiency, but it's mostly about saving time of the employees so they can go
02:48and spend time on more strategic projects.
02:50But really, ROI is measured two ways, efficiencies of the existing employees and new revenue generating opportunities measured in dollars.
02:58Today, we heard that the government is considering, once again, limiting the amount of chips that NVIDIA can sell to
03:04Chinese.
03:05You saw the spat between the government and Anthropik in the U.S. here.
03:09How important is it for Europe to have its own AI companies or each region to have its own strengths
03:17in AI?
03:17Yeah, I think where the world is moving towards is definitely every company, every region becoming more independent in their
03:26usage of AI.
03:28So for us at Miss Scholl, our value proposition is all around, as I mentioned, portability, deploying the technology in
03:34your own roles and customization.
03:36So whether you're looking at a country or at a business in the U.S. or outside of the U
03:42.S., those companies need to be able to rely on their own technology and build their own AI to be
03:47successful.
Comments