The Race for Creating Efficient AI Models while Lowering Carbon Footprint
Ask about Knowledge at Wharton
Episode Description
ABOUT THE EPISODE Wharton professor Stefano Puntoni joins Dan Loney to discuss how AI platforms balance free and paid models, the role of enterprise vs. consumer markets, and the push toward advertising through sponsored bots. He also explores Google’s findings on Gemini’s shrinking carbon footprint and the race to build not just bigger, but more efficient AI systems. The conversation highlights how monetization, trust, and sustainability are shaping the future of artificial intelligence.
ABOUT THE PODCAST This Week in Business features interviews with Wharton faculty about the latest news, fascinating trends, and issues impacting both consumers and the business world. Episodes are recorded at the Wharton School and published twice per week.
#AIResearch #AIEthics #ResponsibleAI #EthicalAI #AIChatbots ----- The Wharton School of the University of Pennsylvania is committed to sharing its intellectual capital through Knowledge at Wharton, the school’s online business analysis journal.
Launched in 1999, Knowledge at Wharton offers free access to:
- Articles, podcasts, and videos highlighting Wharton faculty research and analysis of current business trends - Interviews with book authors from Wharton School Press - In-depth series of curated content like Wharton Executive Education’s Nano Tools for Leaders - A searchable database of more than 10,000 articles covering all aspects of business - A weekly newsletter that delivers Knowledge at Wharton insights directly to your inbox
Visit the Knowledge at Wharton homepage: https://whr.tn/3qUpSBE Subscribe to the Knowledge at Wharton YouTube channel: https://whr.tn/3DEAgAU
Brief Summary
Stephano Pantoni discusses AI monetization, consumer trust, and energy efficiency in AI technology.
Key points
- The Future of AI Monetization. Exploring how AI companies will balance free and paid models to generate revenue.
- AI's Environmental Impact. Google's Gemini chatbot shows significant improvements in carbon footprint efficiency.
Episode highlights
- The Future of AI MonetizationExploring how AI companies will balance free and paid models to generate revenue.“Will companies need both free and paid models?”3:15Sep 3, 2025
- AI's Environmental ImpactGoogle's Gemini chatbot shows significant improvements in carbon footprint efficiency.“A 44-fold decrease in carbon footprint for Gemini is enormous!”7:36Sep 3, 2025
Transcript
There are so many aspects of AI and the AI story still to be written. And to that end, we're going to spend time uh every month on our first podcast of the month talking about various aspects of the AI technology and its connection to us. Our guest will be Stephano Pantoni, marketing professor here at the Wharton School and also co-director of the Wharton Human AI research program. Stephano, great to talk to you. How are you today, sir? >> I'm great. Great to be here. excited about this new segment. >> Thank you, sir. Yeah, it's this is obviously so important uh in terms of uh what we're seeing going on in our lives.
A couple of areas that we wanted to focus on today. First, how AI and platforms will be adjusting to the needs of the consumer. And one of the things I know you're focused on is the potential of monetization of AI platforms. give us a little uh explanation as to what we might be looking forward to >> in a segment like this one. Trying to dissect uh the news and what they mean uh in this field of AI is tricky because there are so many there's so many things happening all the time. But uh to you know pick a couple of things that might be interesting to our audience. one that uh I think has come to the four uh increasingly in the coming uh weeks and
months has been this conversation about uh the monetization of chat bots. And if you look at all large AI uh labs, they have basically pursued so far what you could call a premium strategy where people can sign up and make an account for Chach PT or you know Gemini or Antropica or whatever. And you can do that for free basically with minimal hurdles. Just leave your email account you confirm the email and you're in business. And then paired with that you have a premium subscriber strategy where you would have say for example Chach Plus which would set you off about 20 bucks a month or GP Chach PT Pro which
is uh the more expensive one which is I believe somewhere around a couple hundred a month and so so far the AI labs have been able to attract very vast audiences. If you look at the subscriber base for chacht is enormous but a very tiny percentage of the people using chacht are paying for it. And uh while for the eyabs uh that large user base is crucial in improving the models as you get increasing amount of feedback from users you can make those models better. But at the same time uh pressure will mount on AI companies to uh uh create revenues. Up to now has been uh uh you know enormous amount of private capital uh entering the space. So these
companies have been very well funded without having to worry too much about uh revenues. But that may change over time as uh the industry matures and maybe some of the froth um kind of like starts to fade. And so basically companies will have to think about these free products that they're offering. How am I going to we're going to make some money out of that? >> But will it be necessary for companies to have both versions of that concept of having a free model and a paid model as we move forward? And and how should the consumer think about the fact that here's another kind of technology that
is being monetized and and really forcing them to pay for its use? In many cases >> many many people believe that uh most of the money in the uh AI space is going to be made out of enterprise solution more than consumer solution. So one pathway for AI companies would be that of focusing on the enterprise market. So you could imagine OpenAI investing much more on the uh let's say building a um you know B2B organization selling solutions to corporates and kind of uh um play down or even wind down the consumer part if that turn out to not generate the kind of revenues that they need. Companies like Microsoft which is
very much focused on the corporate market will be focused there. But many also make the argument and I've heard that argument made by people at AI labs too that the consumer market is important because it creates a sight guest and the interest in the technology that broadly people in society are aware and excited about. That's important for a variety of reasons including regulation, investor relations and all that. But also maybe more importantly by having millions and millions of people interacting with the system companies can uh use the data that they gather through those interactions to make the systems better. So a company like OpenAI
probably should be expected to pursue a hybrid strategy where they do pursue the B2B market but they also keep uh pursuing the B2C market. But the question is how do I make money out of the B2C market? And typically what has happened with digital products over the last two three decades that has meant advertising. >> One of the other areas I know you're focused on is that of energy supply and that is an area uh which obviously is drawing a lot of attention right now because of the expected buildout that we're going to have to see in order to have enough energy. But there is a component of that that obviously will
play to the consumer as well. >> Yeah. So we'll uh let's uh let me final make a final point about the monetization part and then I I'll come to this energy thing because there's been a recent newspaper that is quite interesting in that space. Now if uh companies will be under pressure to uh raise revenue through advertising then that means that uh we will be seeing the arrival of sponsored bots. Basically, this would be bots that would be paid by advertisers to promote particular brands. And you can do that in a variety of ways, but we'll see moves in that direction. A couple of weeks ago, Elon Musk, for example, announced that Grock will be uh selling advertising services.
has been s sort of kind of weak we vague into how they exactly going to do that but quoting from the financial times said if a user is typing to solve a problem by asking rock then advertising the specific solution would be ideal at that point in time so you see potentially a value of a chatbot as an interface for advertisers to reach audiences the interesting spin on that would be from the point of the consumer h do I trust a bot if I know that the answers might be sponsored and how do we navigate that trust issue as an AI lab and potentially also regulators might be concerned about disclosure and deception. And so I think this idea of advertising as a revenue engine for the
free product of chatbots and AI companies I think brings up a lot of really interesting and complicated questions around consumer welfare trust in AI companies and regulation. Now moving to uh the uh energy question you were asking. Um just about a week ago um researchers at Google put out a new paper where they uh reported a very uh detailed in-depth analysis of the carbon footprint of uh their Gemini chatbot. And there is also a quite stunning in that uh uh we've seen a tremendous improvement in the efficiency of these models. And so within a 12 months period from May 24 to May 25, Google documented a uh uh 44 fold decrease in car footprint for for Aquery on Gemini and
that is enormous within a one one short year. So climate concerns are real when it comes to AI. We're building gigantic data centers that are going to suck an enormous amount of energy. But those are more uh function of increasing huges of AI than AI being very power hungry. In fact, uh Aquarium on Gemini now already today is equivalent to watching TV for 9 minutes sorry 9 seconds and about uh five drops water consumption. So we we're talking about fast shrinking carbon footprints which like I said you know if you grow a,000% a year even if you're becoming more efficient you're still using a lot more energy but I think for the future we are at least seeing great improvements in a way there
are two races in AI there's a race at the top to build the biggest most performing models and that's what the race that takes all the attention but there's another race going on a little bit under the radar which is that to create the most efficient cheapest model that enable you to have the same outcome comes with a lower carbon footprint. >> Stephano, great to talk to you as always. Thanks very much for your time today. All the best. >> Thank you, Dan. >> You got it. Stephano Pantoni, who is a marketing professor here at the Wharton School and also co-director of the Wharton Human AI research program.





