WHEN: Today, Tuesday, October 1
WHERE: CNBC Evolve: AI Opportunity
Following are excerpts from the unofficial transcripts of panels from CNBC Evolve: AI Opportunity which took place today, Tuesday, October 1 in New York.
Mandatory credit: CNBC Evolve: AI Opportunity.
The NFL’s AI Playbook panel featuring Jennifer Langton, National Football League Advisor, Player Health & Safety Innovation
JENNIFER LANGTON ON COLLECTING HEAD IMPACT DATA DURING GAMES
Langton: We have created a helmet detection system. So, what the computers are now able to do is label and identify every helmet on the field with the data that they are receiving from this chip, in this pad, they are able to locate the player to that helmet. That is a model that now today we are able to give head impacts to every team. Every team now gets the number of head impacts per player per game.
Tyler Mathisen: And what is measuring the head impact? You said it was just RFID to map the helmet to the player. What is measuring the impact?
Langton: Two algorithms. The first algorithm, which we crowd sources with AWS, the first algorithm identifies and labels the helmet, and the second algorithm takes the actual location from the chip in the shoulder pads. Merging those two, what you are able to do is have then a helmet detection system linking those two. So, the frame of the video is then linked to the coordinates on the field to get a head impact per player. Before we started this program, it used to take four days to count head impact in one game. Then we do a blind review. The technology now does it real time by taking the labels – so what computer vision can do with film, identifying it and synchronizing it to data being captured and collected by the player to then have insights to provide us, in this case it is helmet impacts.
Meeting the Future of Media panel featuring Mark Douglas, MNTN President & CEO and Cristóbal Valenzuela, Runway Co-founder & CEO
CRISTÓBAL VALENZUELA ON NEW TECHNOLOGY
Most of the things we have been able to do over the last couple of years and months are things that are completely new. So first, there is a sense of like well, we need to experience it, we need to understand it, we need to play with it. If you haven’t generated video, well, now is the time. If you really want to understand how this works and what opportunities it gives you and what challenges you have just use it. That’s the first thing.
CRISTÓBAL VALENZUELA ON MOVIE TECHNOLOGY
From a technology research perspective, you should expect models to get much better at – and quality and resolutions. But also more importantly, at control. Like, how do you make good movies? It’s not about prompting something. There’s way more aspects of making good stuff. And so over the next, I would say, six to 12 months, you’ll see another function, like big jump in both quality and control.
MARK DOUGLAS ON QUALITY
We’re partnering with companies like Runway and others on various parts of generative AI models. So it generates the idea for the commercial, the spoken script, the actual, you know, speaking the script, the audio, the music and the video. And at this moment in time, the quality is kind of in that order.
MARK DOUGLAS ON VOICE SELLING
People selling their voice, in a sense, maybe less their likeness, but their voice for something is coming pretty rapidly.
MARK DOUGLAS ON INVESTING
Instead of it being viewed as kind of like massive advance, the whole thing is a big advance, but in terms of the application of it, I think it’s like kind of taking, you know, bringing the lowest, you know – people investing the least, and people investing the most and kind of bringing them closer together.
AI & The Law: Big Data, Big Changes panel featuring Katherine B. Forrest, Paul, Weiss, Rifkind, Wharton & Garrison LLP, Digital Technology Group Chair & Partner
KATHERINE B. FORREST ON REAL DANGERS
AI has the potential for such tremendous good that’s being used for a lot of good, but it also has some real dangers that come with it. And one of the ways in which AI can be dangerous is with a kind of asymmetrical capability for issues relating to conventional weapons, or as we call it, CBRN, chemical, biological, radiological or nuclear weaponry, and the ability to put into a non-state actor’s hands, for instance, the capability to do all kinds of mischief. So I can see AI having some issues with creating a kind of instability but I also know that a lot of companies are working hard on trying to keep things safe.
KATHERINE B. FORREST ON COMMON LAW
In terms of the U.S. setting sort of the stage for things, we’re behind in terms of worldwide regulation but the United States, when people ask me, is AI actually regulated? It’s not regulated is it in the United States, I always say, look we have the common law. The common law applied 200 years ago, and it continues to apply. It’s a very flexible doctrine. You can use the common law to apply to all kinds of incidents that could occur with regard to AI.
KATHERINE B. FORREST ON REGULATION
One of the issues really right now is that if regulation becomes so stringent that it is difficult for companies to comply, or they choose not to stay in business in the United States, we may see companies choosing to put their AI training elsewhere, to put their AI processing elsewhere, to put their AI development elsewhere, because, you know, the regulatory environment could become tough. We just don’t know yet.
KATHERINE B. FORREST ON THE HYPE OF AI
AI is not hype. It’s not going to affect every single business in the same way. So people who were waiting for AI to monolithically transform American business in the same way and are disappointed because it hasn’t absolutely transformed their world, calm down. It’s going to get to you and what you need to do though is assess what’s the competition doing. Don’t get left behind. Make sure that you and your business have the right structures in place to be able to put AI into the business, be able to put compliance regimes around that, that you are able to then understand what the global framework is for your business if you are seeking to go global, because it won’t be the same outside of the U.S. as inside of the U.S. Make sure your board is informed. You know, people do not want to forget that the leadership of a company needs to be made aware and kept aware of how AI is being used, ensure that you’ve got a common definition of AI across the company. One thing I see frequently is there’s about 27 different definitions of AI in a company with about 15 different units, and that’s a problem because everybody then is regulating to a different definition. So I would say to an American company, or to any company, keep up with what’s happening. This is the most transformative moment.
The Reality of AI in Real Estate with Ryan Serhant panel featuring Ryan Serhant, SERHANT CEO; “Owning Manhattan” Star
RYAN SERHANT ON INCREASED CLOSING COST
It’s the same thing with taxes. If you’re going to increase closing cost in a place like New York City or anywhere, someone’s going to pay it and it’s either going to be the buyer or the seller, it’s like minimum wage. Someone’s going to pay it, and it’s usually on the consumer even though they’re the ones who voted for it.
RYAN SERHANT ON CUTTING RATES
The monthly cost of every one hundred thousand dollars borrowed right now with let’s say a half a point interest swing is 33 dollars. So people sit and they’ll wait and say rates went up half a point I can’t do anything, or rates came down half a point we’re going to run into the market. Everyone says that’s a material change, it’s not.
RYAN SERHANT ON THE HOUSING MARKET
Everyone’s paying attention to what’s going to happen in the fall, the Fed’s going to cut rates once, twice, we don’t know. What’s going to happen with the election? You know what, we don’t know what to do so we’re just going to hold to the detriment of the entire market. But smart people are jumping in and saying, you know what I can marry the house, I’ll date the rate and I’ll get what I want and negotiate.
AI’s Open-Source Future panel featuring Clément Delangue, Hugging Face Co-Founder & CEO
CLÉMENT DELANGUE ON AI CREATING HEALTHY COMPETITION AMONG COMPANIES
I think it creates healthy competition. AI is kind of very foundational technology and I think you don’t want it only in the hands of a few companies. Like, imagine if you had a world where only a few companies were able to do software. It would kind of be like a scary world. I think open source comes in as a way to create more competition to give more organizations and more companies the power to also build AI, build their own systems that they control to make sure they don’t only rely on the technology companies.
CLÉMENT DELANGUE ON OPEN-SOURCE
When we were trying to build a use case for a new company, specialized use case, now the open-source models are better than the proprietary models. You only want to use proprietary models most of the time for generic use case when you’re doing search, when you’re doing ChatGPT, which is your very general use case. Most of the other domains, not only in texts but also in image, video, in biology, chemistry, now amazing foundations in open source for you to really build AI.
CLÉMENT DELANGUE ON AI LIFTING ALL BOATS
We’re trying – to work with them so that they can keep contributing to the field, so a lot of them are sharing, actually, open source models, open data sets, applications under the Hugging Face platform. And when you talk to them, you realize they are all interested in the progress of the field, right. So in a way, open source AI is the tide that lifts all boats.
CLÉMENT DELANGUE ON AI TURNOVER
Even the biggest AI companies still are facing questions about the sustainability of their revenue, of their business models. So I think a lot of teams right now in AI are struggling. It also creates a lot of pressure for founders or leadership teams at AI companies. So I think it leads to a little bit of this shuffle and this kind of faster turnover within the AI companies.
CLÉMENT DELANGUE ON AI REGULATION
What works well for foundational technology like that, similarly to what has been done in software, is not to regulate the science, not to regulate the infrastructure, but regulate can affect the final use cases, right. Making sure that when AI is applied to the use case, it’s applied for good.
For more information contact:
Jessica Iacobazzo
About CNBC
CNBC is the recognized world leader in business news, providing real-time financial market coverage and business content consumed by nearly half a billion people per month across all platforms. The network’s 14 live hours a day of news programming in North America (weekdays from 5:00 a.m. – 7:00 p.m. ET) is produced at CNBC’s global headquarters in Englewood Cliffs, N.J., and includes reports from CNBC News bureaus worldwide. CNBC at night features a mix of reality programming, CNBC’s highly successful series produced exclusively for CNBC and a number of distinctive in-house documentaries.
CNBC also offers content through its vast portfolio of digital products such as: CNBC.com, which provides financial market news and information to CNBC’s investor audience; CNBC Make It, a digital destination focused on making you smarter about how you earn, save and spend your money; CNBC PRO, a premium service that provides in-depth access to Wall Street; a suite of CNBC mobile apps for iOS and Android devices; Amazon Alexa, Google Assistant and Apple Siri voice interfaces; and streaming services including Apple TV, Roku, Amazon Fire TV, Android TV and Samsung Smart TVs. To learn more, visit https://www.cnbc.com/digital-products/.
Members of the media can receive more information about CNBC and its programming on the NBCUniversal Media Village Web site at http://www.nbcumv.com/programming/cnbc. For more information about NBCUniversal, please visit http://www.NBCUniversal.com.
