At its Build conference this week, Microsoft (MSFT) unveiled the Copilot+ PC, designed to integrate artificial intelligence capabilities into its computers. RBC Capital Markets software equity analyst Rishi Jaluria and Macquarie's head of U.S. AI and software research Frederick Havemeyer appear on Market Domination to discuss the company's latest AI strategy.
“I think the introduction of these AI PCs is really going to be a game changer in putting generative AI in the hands of every knowledge worker out there,” Jaluria explained. He added that users will be able to use these small to medium language models locally, increasing the use of AI for general purpose purposes. Regarding larger language models, Havemeyer said, “It's very clear to us that the technology for generative AI models for large language models is something that will be consumer-ready over time. ChatGPT was, and I think that's right. The interface and the right platform will be found over time.”
AI is permeating every sector as businesses look to innovate and gain efficiencies. “80% of CIOs say they are currently adopting AI or plan to adopt AI in the next 12 months, so the demand is real and the use cases are real,” Jaluria explains, adding that there's still a lot of room for the technology to grow.
For more expert insights and the latest market trends, click here to watch this full episode of Market Domination.
This post was written by Melanie Leal
Video Transcript
Microsoft is committed to AI.
The company this week announced a new category of PC CS called Copilot and computers equipped with so-called AI PC chips, which will now run on the Open A GP T +40 model.
As software technology continues to advance, we’re looking at how to navigate the bigger picture and how best to position your portfolio in the Yahoo Finance playbook.
Let's welcome Rishi Julia R, software equity analyst, BC Capital Markets, and Frederick Havemeyer, head of USA I and software research at Macquarie.
Gentlemen.
Thank you very much for being here. I want to start with Microsoft's announcement itself, how you should think about it as an investor, how big it will be, how incremental it will be to your revenue, etc. Let's start with you.
Yes, absolutely.
And thank you so much for inviting us.
We were just at Microsoft Build in Seattle this week, and we think the introduction of these AI PCs is going to be a real game changer for putting generative AI in the hands of every knowledge worker.
right.
Remember that using any type of GP T system today requires a high-speed Internet connection and a large amount of computing power.
As you know, it will take some time before it becomes lightweight enough for everyone to use, but you can use some of these small and medium language models locally without having a super fast internet connection. It will look like this.
I think doing so allows for more versatility.
Well, AI use cases and workloads that didn't exist before.
So I think this is a very exciting announcement for Microsoft, Fred.
Let's take him here too.
Same question, Fred.
So you know your reaction to that Microsoft event, Rishi says these new A IP CS are a game changer, Fred, do you agree?
So over the past year, we've consistently said that AI needs to move to the edge and move to devices.
And in many cases, that requires the kind of hardware that Microsoft is deploying.
And this is what excites us both about being successful for both consumers and knowledge workers, and being able to accelerate the adoption of AI in a profitable way.
If you recall, last time I was at Yahoo Finance, we had a fun time talking about what smaller language models mean, and what larger language models, which are still pretty big, but still smaller, mean for margins.
Based on our past research, we believe that running generative AI models to achieve 80%-90% plus gross margins can be achieved using small language models on dedicated hardware or local devices. .
So, I think that's very interesting, Fred, follow up on that, you wrote a note recently about how consumers judge the impact here.
And we continue to try to discern whether where we are in the AI cycle is still a push cycle or whether we're starting to become a pull cycle.
In other words, are these products coming out because retail customers specifically want them, or are they in situations where companies have to convince customers that they want them? is.
I think when you look at a product like Chat G BT, we already know that there are over 100 million users who often qualify as consumers who are really interested in these types of products.
Now we are commercializing it and making it convenient and interesting for consumers.
There is probably a technology out there that you have invented to solve a problem.
At the same time, we think it's very clear to us that the technology of large-scale language modeling generative AI models is going to be consumer-adaptive over time.
Chat GP T did it.
And I think that with the right interface and the right platform, you'll find that over time.
So, I'm a user of Meta's Ray-Ban glasses and I'm only here to discuss the glasses, not to discuss Meta.
That's interesting.
These are really cool and it's really fun to have these AI models sitting inside my glasses and doing this kind of work.
So, let's take some time and look.
Richie.
Same question.
So, Richie, as you look across your coverage, what kind of customer adoption and engagement are you seeing around these new AI products and capabilities?
Yeah.
Hmm, of course.
And I'm glad to have Fred with me because I definitely agree with a lot of what he's saying.
You see, um, what I'm trying to say is, when we're talking to businesses, we're talking to the head of Cio S CTO S.
Enthusiasm for using AI is growing within the enterprise.
Well, a lot of it, I would say probably 70% is actual net new budget.
So it doesn't even cannibalize other parts.
It comes from other parts of the organization.
Well, money is actually allocated.
Um, we say, um, 80% of CIOs are currently implementing AI or plan to implement AI within the next 12 months.
In other words, appetite is real.
The usage examples are real.
And remember, we're still in the very early stages of this effort.
If we compare it to the Internet stage, the GP T 40 is incredibly powerful, but it's probably at the 14 4 dial-up level, right?
We haven't even reached high-speed broadband yet.
So I think there's still a lot of room for this technology to advance from here, but we're already seeing real interest from businesses and consumers as well.
