AI will touch ‘every part of our lives, every form of human labor,’ analyst explains

ARK Invest Co-Lead, ARK Venture and Analyst Will Summerlin discusses the rise of ChatGPT and AI coding assistants and details the potential value creation of AI technology.

影片文字轉錄稿

[AUDIO LOGO]

- This last quarter saw the rise of artificial intelligence, AI, with ChatGPT launching the technology into the spotlight. But with many of tech's biggest players working on AI products of their own, who are the winners, and what's the investment opportunity there? To discuss this, we want to bring in Will Summerlin. He's ARK Invest Co-Lead and Analyst.

Will, it's great to see you again. So during this earnings season, our colleague Josh Schafer was just saying that, of the big tech companies, AI was mentioned over 100 times on the earnings call. What was your big takeaway from all this hype surrounding AI?

WILL SUMMERLIN: Well, I think people are realizing its potential. With ChatGPT, we saw the fastest-growing software product in history. It grew from 0 to 100 billion users in 60 days. And so, this isn't hype. This is customers seeing real value out of technology.

I think when you think about the investable universe of AI, we really think about it in two buckets. On one hand, we have the infrastructure plays, the picks and shovels plays. So think of NVIDIA, for example, on the public side. On the private side, we just invested in a company called MosaicML, which helps companies train their own AI models.

And then the other bucket is applications. So companies that are actually embracing AI and building customer experiences around that, helping customers solve problems using AI. ChatGPT is an example of an application. Another example of an application is AI coding assistance. There is a company called Replit, which is a growth-stage startup that has a chat assistant or an AI coding assistant called Ghostwriter that's blown up in popularity. It's making software engineers more than twice as productive.

And if you think about the value creation opportunity here, knowledge workers make about $32 trillion a year, collectively. And we're already seeing cases where AI is doubling their productivity. And so, the value creation here is going to be incredible. We think it might actually be the greatest form of value creation we've seen out of any technology trend-- greater than the internet, greater than mobile, greater than cloud. And so there's certainly excitement around the opportunity, and companies are really pivoting to capture the potential.

- That is quite a significant statement, Will. In terms of sectors you see poised to benefit the most, is there one that stands out?

WILL SUMMERLIN: Yeah. I think knowledge work, broadly. This applies to legal. This applies to software engineering. This applies to most domains of knowledge work. The incredible thing about AI is it really automates thinking.

We think about technology throughout history, it's automated physical labor. And so we saw this with a tractor. It made farms 17 times as productive. And we're starting to see this with knowledge work, where AI is actually helping us do tasks that require a lot of thinking.

And so software engineering is probably the best example we've seen so far. We're seeing applications like Replit Ghostwriter. We're also seeing this from GitHub Copilot. More than double the productivity of software engineers, and that's a class of knowledge workers that's highly paid, $200,000-a-year software engineer that gets double the productivity all of a sudden because of an AI coding assistant. That's incredible.

We're also seeing this in legal, right? We're seeing assistance in legal that reduce the cost of contract reviews by 70%, 80%, 90%, because instead of having a human go through every single line in the contract, you can use AI to do the review, and then a human just to look at the things where there may be issues. And so we think this is going to apply broadly.

There are also areas where there is significant cost-savings potential. And so if we think about contact centers, for example. Labor in contact centers is challenging. It's a relatively unsatisfactory job, and so the churn is very high. In many cases, the churn is over 100%, meaning people actually don't stay an entire year once they join a contact center staff. So you train an employee, and then three, four, or five months later, they leave.

That's an area where it's ripe for automation. We're starting to see companies like Twilio embrace that opportunity and build AI models to really automate contact centers. And so I think this is going to touch pretty much every form of our life or every part of our lives and pretty much every form of human labor.

- So Will, it sounds like you think it's going to have a significant impact on the job market. What's the extent of that?

WILL SUMMERLIN: If you look at technology throughout history, everyone always says that technology is going to lead to mass unemployment, and that's not been true in the past. Technology has always led to a higher quality of life, and it's actually spurred job growth.

And I think the same thing is going to happen here. Humans are creative beings, and when we're able to get leverage over a time with technology, we find ways to create value. And so I think, ultimately, this will actually lead to a higher quality of life, and probably be an accelerate to job growth.

- Will, you've got some winners, and we have to ask you about Tesla and their advantage in this space.

WILL SUMMERLIN: So when we think about investing in application layer AI companies, we think about it on three axes. The first is a data advantage. Can they develop a data advantage that allows them to train superior AI models and really outcompete their competitors? That's an advantage that grows with scale.

And so like network effects, the more data you have, the better your models become. The better your models, the more customers you get. The more customers you get, the more data you get. And this flywheel sort of continues. We've seen that with Tesla. They have billions of miles recorded through their full self-driving suite, and that's going to be hard for others to compete with. So they have a data advantage that's continuing to scale.

Second is vision. Do they have a leader at the company who is going all-in on AI and burning the boats and committing the company to the future of AI? And Elon is definitely doing that. It would be hard to argue that he's not. And third is distribution. Do they have a distribution advantage that they can leverage to deploy AI capabilities? Tesla certainly does. They already have a large fleet of vehicles, and the AI autopilot or full self-driving capability is included as a feature in those vehicles. So it certainly meets the three criteria for us

- Well, Will, you're certainly making a compelling case for that. Will Summerlin, great to have you. Thanks so much.